thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-27 01:51:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' of github.com:yandex/ClickHouse

Browse Source
This commit is contained in:
BayoNet 2019-04-24 11:31:52 +03:00
commit eb8df2978a
103 changed files with 1167 additions and 2686 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
cmake/find_execinfo.cmake
View File

@ -1,9 +1,8 @@
if (OS_FREEBSD)
find_library (EXECINFO_LIBRARY execinfo)
find_library (ELF_LIBRARY elf)
message (STATUS "Using execinfo: ${EXECINFO_LIBRARY}")
message (STATUS "Using elf: ${ELF_LIBRARY}")
set (EXECINFO_LIBRARIES ${EXECINFO_LIBRARY} ${ELF_LIBRARY})
message (STATUS "Using execinfo: ${EXECINFO_LIBRARIES}")
else ()
set (EXECINFO_LIBRARY "")
set (ELF_LIBRARY "")
set (EXECINFO_LIBRARIES "")
endif ()

2
contrib/lz4 vendored

@ -1 +1 @@
Subproject commit c10863b98e1503af90616ae99725ecd120265dfb
Subproject commit 780aac520b69d6369f4e3995624c37e56d75498d

3
contrib/lz4-cmake/CMakeLists.txt
View File

@ -9,8 +9,7 @@ add_library (lz4
${LIBRARY_DIR}/xxhash.h
${LIBRARY_DIR}/lz4.h
${LIBRARY_DIR}/lz4hc.h
${LIBRARY_DIR}/lz4opt.h)
${LIBRARY_DIR}/lz4hc.h)
target_compile_definitions(lz4 PUBLIC LZ4_DISABLE_DEPRECATE_WARNINGS=1)

3
dbms/CMakeLists.txt
View File

@ -194,8 +194,7 @@ target_link_libraries (clickhouse_common_io
${CITYHASH_LIBRARIES}
PRIVATE
${ZLIB_LIBRARIES}
${EXECINFO_LIBRARY}
${ELF_LIBRARY}
${EXECINFO_LIBRARIES}
PUBLIC
${Boost_SYSTEM_LIBRARY}
PRIVATE

2
dbms/programs/clang/Compiler-5.0.0/CMakeLists.txt
View File

@ -46,7 +46,7 @@ LLVMSupport
#PollyISL
#PollyPPCG
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARY} Threads::Threads
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARIES} Threads::Threads
${MALLOC_LIBRARIES}
${GLIBC_COMPATIBILITY_LIBRARIES}
${MEMCPY_LIBRARIES}

2
dbms/programs/clang/Compiler-6.0.0/CMakeLists.txt
View File

@ -46,7 +46,7 @@ ${REQUIRED_LLVM_LIBRARIES}
#PollyISL
#PollyPPCG
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARY} Threads::Threads
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARIES} Threads::Threads
${MALLOC_LIBRARIES}
${GLIBC_COMPATIBILITY_LIBRARIES}
${MEMCPY_LIBRARIES}

2
dbms/programs/clang/Compiler-7.0.0/CMakeLists.txt
View File

@ -42,7 +42,7 @@ lldCore
${REQUIRED_LLVM_LIBRARIES}
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARY} Threads::Threads
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARIES} Threads::Threads
${MALLOC_LIBRARIES}
${GLIBC_COMPATIBILITY_LIBRARIES}
${MEMCPY_LIBRARIES}

2
dbms/programs/clang/Compiler-7.0.0bundled/CMakeLists.txt
View File

@ -42,7 +42,7 @@ lldCore
${REQUIRED_LLVM_LIBRARIES}
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARY} Threads::Threads
PUBLIC ${ZLIB_LIBRARIES} ${EXECINFO_LIBRARIES} Threads::Threads
${MALLOC_LIBRARIES}
${GLIBC_COMPATIBILITY_LIBRARIES}
${MEMCPY_LIBRARIES}

2
dbms/programs/copier/ClusterCopier.cpp
View File

@ -2054,7 +2054,7 @@ private:
ConfigurationPtr task_cluster_initial_config;
ConfigurationPtr task_cluster_current_config;
Coordination::Stat task_descprtion_current_stat;
Coordination::Stat task_descprtion_current_stat{};
std::unique_ptr<TaskCluster> task_cluster;

2
dbms/programs/performance-test/ConfigPreprocessor.cpp
View File

@ -16,7 +16,7 @@ std::vector<XMLConfigurationPtr> ConfigPreprocessor::processConfig(
std::vector<XMLConfigurationPtr> result;
for (const auto & path : paths)
{
result.emplace_back(new XMLConfiguration(path));
result.emplace_back(XMLConfigurationPtr(new XMLConfiguration(path)));
result.back()->setString("path", Poco::Path(path).absolute().toString());
}

2
dbms/programs/performance-test/PerformanceTestSuite.cpp
View File

@ -298,6 +298,8 @@ std::unordered_map<std::string, std::vector<std::size_t>> getTestQueryIndexes(co
{
std::unordered_map<std::string, std::vector<std::size_t>> result;
const auto & options = parsed_opts.options;
if (options.empty())
return result;
for (size_t i = 0; i < options.size() - 1; ++i)
{
const auto & opt = options[i];

262
dbms/scripts/gen-bias-data.py
View File

@ -1,262 +0,0 @@
#!/usr/bin/python3
# -*- coding: utf-8 -*-
import sys
import argparse
import tempfile
import random
import subprocess
import bisect
from copy import deepcopy
# Псевдослучайный генератор уникальных чисел.
# http://preshing.com/20121224/how-to-generate-a-sequence-of-unique-random-integers/
class UniqueRandomGenerator:
prime = 4294967291
def __init__(self, seed_base, seed_offset):
self.index = self.permutePQR(self.permutePQR(seed_base) + 0x682f0161)
self.intermediate_offset = self.permutePQR(self.permutePQR(seed_offset) + 0x46790905)
def next(self):
val = self.permutePQR((self.permutePQR(self.index) + self.intermediate_offset) ^ 0x5bf03635)
self.index = self.index + 1
return val
def permutePQR(self, x):
if x >=self.prime:
return x
else:
residue = (x * x) % self.prime
if x <= self.prime/2:
return residue
else:
return self.prime - residue
# Создать таблицу содержащую уникальные значения.
def generate_data_source(host, port, http_port, min_cardinality, max_cardinality, count):
chunk_size = round((max_cardinality - min_cardinality) / float(count))
used_values = 0
cur_count = 0
next_size = 0
sup = 32768
n1 = random.randrange(0, sup)
n2 = random.randrange(0, sup)
urng = UniqueRandomGenerator(n1, n2)
is_first = True
with tempfile.TemporaryDirectory() as tmp_dir:
filename = tmp_dir + '/table.txt'
with open(filename, 'w+b') as file_handle:
while cur_count < count:
if is_first == True:
is_first = False
if min_cardinality != 0:
next_size = min_cardinality + 1
else:
next_size = chunk_size
else:
next_size += chunk_size
while used_values < next_size:
h = urng.next()
used_values = used_values + 1
out = str(h) + "\t" + str(cur_count) + "\n";
file_handle.write(bytes(out, 'UTF-8'));
cur_count = cur_count + 1
query = "DROP TABLE IF EXISTS data_source"
subprocess.check_output(["clickhouse-client", "--host", host, "--port", str(port), "--query", query])
query = "CREATE TABLE data_source(UserID UInt64, KeyID UInt64) ENGINE=TinyLog"
subprocess.check_output(["clickhouse-client", "--host", host, "--port", str(port), "--query", query])
cat = subprocess.Popen(("cat", filename), stdout=subprocess.PIPE)
subprocess.check_output(("POST", "http://{0}:{1}/?query=INSERT INTO data_source FORMAT TabSeparated".format(host, http_port)), stdin=cat.stdout)
cat.wait()
def perform_query(host, port):
query = "SELECT runningAccumulate(uniqExactState(UserID)) AS exact, "
query += "runningAccumulate(uniqCombinedRawState(UserID)) AS approx "
query += "FROM data_source GROUP BY KeyID"
return subprocess.check_output(["clickhouse-client", "--host", host, "--port", port, "--query", query])
def parse_clickhouse_response(response):
parsed = []
lines = response.decode().split("\n")
for cur_line in lines:
rows = cur_line.split("\t")
if len(rows) == 2:
parsed.append([float(rows[0]), float(rows[1])])
return parsed
def accumulate_data(accumulated_data, data):
if not accumulated_data:
accumulated_data = deepcopy(data)
else:
for row1, row2 in zip(accumulated_data, data):
row1[1] += row2[1];
return accumulated_data
def generate_raw_result(accumulated_data, count):
expected_tab = []
bias_tab = []
for row in accumulated_data:
exact = row[0]
expected = row[1] / count
bias = expected - exact
expected_tab.append(expected)
bias_tab.append(bias)
return [ expected_tab, bias_tab ]
def generate_sample(raw_estimates, biases, n_samples):
result = []
min_card = raw_estimates[0]
max_card = raw_estimates[len(raw_estimates) - 1]
step = (max_card - min_card) / (n_samples - 1)
for i in range(0, n_samples + 1):
x = min_card + i * step
j = bisect.bisect_left(raw_estimates, x)
if j == len(raw_estimates):
result.append((raw_estimates[j - 1], biases[j - 1]))
elif raw_estimates[j] == x:
result.append((raw_estimates[j], biases[j]))
else:
# Найти 6 ближайших соседей. Вычислить среднее арифметическое.
# 6 точек слева x [j-6 j-5 j-4 j-3 j-2 j-1]
begin = max(j - 6, 0) - 1
end = j - 1
T = []
for k in range(end, begin, -1):
T.append(x - raw_estimates[k])
# 6 точек справа x [j j+1 j+2 j+3 j+4 j+5]
begin = j
end = min(j + 5, len(raw_estimates) - 1) + 1
U = []
for k in range(begin, end):
U.append(raw_estimates[k] - x)
# Сливаем расстояния.
V = []
lim = min(len(T), len(U))
k1 = 0
k2 = 0
while k1 < lim and k2 < lim:
if T[k1] == U[k2]:
V.append(j - k1 - 1)
V.append(j + k2)
k1 = k1 + 1
k2 = k2 + 1
elif T[k1] < U[k2]:
V.append(j - k1 - 1)
k1 = k1 + 1
else:
V.append(j + k2)
k2 = k2 + 1
if k1 < len(T):
while k1 < len(T):
V.append(j - k1 - 1)
k1 = k1 + 1
elif k2 < len(U):
while k2 < len(U):
V.append(j + k2)
k2 = k2 + 1
# Выбираем 6 ближайших точек.
# Вычисляем средние.
begin = 0
end = min(len(V), 6)
sum = 0
bias = 0
for k in range(begin, end):
sum += raw_estimates[V[k]]
bias += biases[V[k]]
sum /= float(end)
bias /= float(end)
result.append((sum, bias))
# Пропустить последовательные результаты, чьи оценки одинаковые.
final_result = []
last = -1
for entry in result:
if entry[0] != last:
final_result.append((entry[0], entry[1]))
last = entry[0]
return final_result
def dump_arrays(data):
print("Size of each array: {0}\n".format(len(data)))
is_first = True
sep = ''
print("raw_estimates = ")
print("{")
for row in data:
print("\t{0}{1}".format(sep, row[0]))
if is_first == True:
is_first = False
sep = ","
print("};")
is_first = True
sep = ""
print("\nbiases = ")
print("{")
for row in data:
print("\t{0}{1}".format(sep, row[1]))
if is_first == True:
is_first = False
sep = ","
print("};")
def start():
parser = argparse.ArgumentParser(description = "Generate bias correction tables for HyperLogLog-based functions.")
parser.add_argument("-x", "--host", default="localhost", help="ClickHouse server host name");
parser.add_argument("-p", "--port", type=int, default=9000, help="ClickHouse server TCP port");
parser.add_argument("-t", "--http_port", type=int, default=8123, help="ClickHouse server HTTP port");
parser.add_argument("-i", "--iterations", type=int, default=5000, help="number of iterations");
parser.add_argument("-m", "--min_cardinality", type=int, default=16384, help="minimal cardinality");
parser.add_argument("-M", "--max_cardinality", type=int, default=655360, help="maximal cardinality");
parser.add_argument("-s", "--samples", type=int, default=200, help="number of sampled values");
args = parser.parse_args()
accumulated_data = []
for i in range(0, args.iterations):
print(i + 1)
sys.stdout.flush()
generate_data_source(args.host, str(args.port), str(args.http_port), args.min_cardinality, args.max_cardinality, 1000)
response = perform_query(args.host, str(args.port))
data = parse_clickhouse_response(response)
accumulated_data = accumulate_data(accumulated_data, data)
result = generate_raw_result(accumulated_data, args.iterations)
sampled_data = generate_sample(result[0], result[1], args.samples)
dump_arrays(sampled_data)
if __name__ == "__main__": start()

1
dbms/scripts/gen_benchmark_data/README.md
View File

@ -1 +0,0 @@
Hits table generator based on LSTM neural network trained on real hits. You need to have weights for model or train model on real hits to generate data.

22
dbms/scripts/gen_benchmark_data/generate.py
View File

@ -1,22 +0,0 @@
import argparse
from model import Model
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('-n', type=int, default=100000,
help='number of objects to generate')
parser.add_argument('--output_file', type=str, default='out.tsv',
help='output file name')
parser.add_argument('--weights_path', type=str,
help='path to weights')
args = parser.parse_args()
if __name__ == '__main__':
if not args.weights_path:
raise Exception('please specify path to model weights with --weights_path')
gen = Model()
gen.generate(args.n, args.output_file, args.weights_path)

147
dbms/scripts/gen_benchmark_data/model.py
View File

@ -1,147 +0,0 @@
import numpy as np
import os
import pickle
import tensorflow as tf
from random import sample
from keras.layers import Dense, Embedding
from tqdm import tqdm
RNN_NUM_UNITS = 256
EMB_SIZE = 32
MAX_LENGTH = 1049
with open('tokens', 'rb') as f:
tokens = pickle.load(f)
n_tokens = len(tokens)
token_to_id = {c: i for i, c in enumerate(tokens)}
def to_matrix(objects, max_len=None, pad=0, dtype='int32'):
max_len = max_len or max(map(len, objects))
matrix = np.zeros([len(objects), max_len], dtype) + pad
for i in range(len(objects)):
name_ix = list(map(token_to_id.get, objects[i]))
matrix[i, :len(name_ix)] = name_ix
return matrix.T
class Model:
def __init__(self, learning_rate=0.0001):
# an embedding layer that converts character ids into embeddings
self.embed_x = Embedding(n_tokens, EMB_SIZE)
get_h_next = Dense(1024, activation='relu')
# a dense layer that maps current hidden state
# to probabilities of characters [h_t+1]->P(x_t+1|h_t+1)
self.get_probas = Dense(n_tokens, activation='softmax')
self.input_sequence = tf.placeholder('int32', (MAX_LENGTH, None))
batch_size = tf.shape(self.input_sequence)[1]
self.gru_cell_first = tf.nn.rnn_cell.GRUCell(RNN_NUM_UNITS)
self.lstm_cell_second = tf.nn.rnn_cell.LSTMCell(RNN_NUM_UNITS)
h_prev_first = self.gru_cell_first.zero_state(batch_size, dtype=tf.float32)
h_prev_second = tf.nn.rnn_cell.LSTMStateTuple(
tf.zeros([batch_size, RNN_NUM_UNITS]), # initial cell state,
tf.zeros([batch_size, RNN_NUM_UNITS]) # initial hidden state
)
predicted_probas = []
for t in range(MAX_LENGTH):
x_t = self.input_sequence[t]
# convert character id into embedding
x_t_emb = self.embed_x(tf.reshape(x_t, [-1, 1]))[:, 0]
out_next_first, h_next_first = self.gru_cell_first(x_t_emb, h_prev_first)
h_prev_first = h_next_first
out_next_second, h_next_second = self.lstm_cell_second(out_next_first, h_prev_second)
h_prev_second = h_next_second
probas_next = self.get_probas(out_next_second)
predicted_probas.append(probas_next)
predicted_probas = tf.stack(predicted_probas)
predictions_matrix = tf.reshape(predicted_probas[:-1], [-1, len(tokens)])
answers_matrix = tf.one_hot(tf.reshape(self.input_sequence[1:], [-1]), n_tokens)
self.loss = tf.reduce_mean(tf.reduce_sum(
-answers_matrix * tf.log(tf.clip_by_value(predictions_matrix, 1e-7, 1.0)),
reduction_indices=[1]
))
optimizer = tf.train.AdamOptimizer(learning_rate)
gvs = optimizer.compute_gradients(self.loss)
capped_gvs = [(gr if gr is None else tf.clip_by_value(gr, -1., 1.), var) for gr, var in gvs]
self.optimize = optimizer.apply_gradients(capped_gvs)
self.sess = tf.Session()
self.sess.run(tf.global_variables_initializer())
self.saver = tf.train.Saver()
def train(self, train_data_path, save_dir, num_iters, batch_size=64, restore_from=False):
history = []
if restore_from:
with open(restore_from + '_history') as f:
history = pickle.load(f)
self.saver.restore(self.sess, restore_from)
with open(train_data_path, 'r') as f:
train_data = f.readlines()
train_data = filter(lambda a: len(a) < MAX_LENGTH, train_data)
for i in tqdm(range(num_iters)):
batch = to_matrix(
map(lambda a: '\n' + a.rstrip('\n'), sample(train_data, batch_size)),
max_len=MAX_LENGTH
)
loss_i, _ = self.sess.run([self.loss, self.optimize], {self.input_sequence: batch})
history.append(loss_i)
if len(history) % 2000 == 0:
self.saver.save(self.sess, os.path.join(save_dir, '{}_iters'.format(len(history))))
self.saver.save(self.sess, os.path.join(save_dir, '{}_iters'.format(len(history))))
with open(os.path.join(save_dir, '{}_iters_history'.format(len(history)))) as f:
pickle.dump(history, f)
def generate(self, num_objects, output_file, weights_path):
self.saver.restore(self.sess, weights_path)
batch_size = num_objects
x_t = tf.placeholder('int32', (None, batch_size))
h_t_first = tf.Variable(tf.zeros([batch_size, RNN_NUM_UNITS]))
h_t_second = tf.nn.rnn_cell.LSTMStateTuple(
tf.Variable(tf.zeros([batch_size, RNN_NUM_UNITS])),
tf.Variable(tf.zeros([batch_size, RNN_NUM_UNITS]))
)
x_t_emb = self.embed_x(tf.reshape(x_t, [-1, 1]))[:, 0]
first_out_next, next_h_first = self.gru_cell_first(x_t_emb, h_t_first)
second_out_next, next_h_second = self.lstm_cell_second(first_out_next, h_t_second)
next_probs = self.get_probas(second_out_next)
x_sequence = np.zeros(shape=(1, batch_size), dtype=int) + token_to_id['\n']
self.sess.run(
[tf.assign(h_t_first, h_t_first.initial_value),
tf.assign(h_t_second[0], h_t_second[0].initial_value),
tf.assign(h_t_second[1], h_t_second[1].initial_value)]
)
for i in tqdm(range(MAX_LENGTH - 1)):
x_probs, _, _, _ = self.sess.run(
[next_probs,
tf.assign(h_t_second[0], next_h_second[0]),
tf.assign(h_t_second[1], next_h_second[1]),
tf.assign(h_t_first, next_h_first)],
{x_t: [x_sequence[-1, :]]}
)
next_char = [np.random.choice(n_tokens, p=x_probs[i]) for i in range(batch_size)]
if sum(next_char) == 0:
break
x_sequence = np.append(x_sequence, [next_char], axis=0)
with open(output_file, 'w') as f:
f.writelines([''.join([tokens[ix] for ix in x_sequence.T[k]]) + '\n' for k in range(batch_size)])

3
dbms/scripts/gen_benchmark_data/requirements.txt
View File

@ -1,3 +0,0 @@
Keras==2.0.6
numpy
tensorflow-gpu==1.4.0

506
dbms/scripts/gen_benchmark_data/tokens
View File

@ -1,506 +0,0 @@
(lp0
S'\x83'
p1
aS'\x04'
p2
aS'\x87'
p3
aS'\x8b'
p4
aS'\x8f'
p5
aS'\x10'
p6
aS'\x93'
p7
aS'\x14'
p8
aS'\x97'
p9
aS'\x18'
p10
aS'\x9b'
p11
aS'\x1c'
p12
aS'\x9f'
p13
aS' '
p14
aS'\xa3'
p15
aS'$'
p16
aS'\xa7'
p17
aS'('
p18
aS'\xab'
p19
aS','
p20
aS'\xaf'
p21
aS'0'
p22
aS'\xb3'
p23
aS'4'
p24
aS'\xb7'
p25
aS'8'
p26
aS'\xbb'
p27
aS'<'
p28
aS'\xbf'
p29
aS'@'
p30
aS'\xc3'
p31
aS'D'
p32
aS'\xc7'
p33
aS'H'
p34
aS'\xcb'
p35
aS'L'
p36
aS'\xcf'
p37
aS'P'
p38
aS'\xd3'
p39
aS'T'
p40
aS'\xd7'
p41
aS'X'
p42
aS'\xdb'
p43
aS'\\'
p44
aS'\xdf'
p45
aS'`'
p46
aS'\xe3'
p47
aS'd'
p48
aS'\xe7'
p49
aS'h'
p50
aS'\xeb'
p51
aS'l'
p52
aS'\xef'
p53
aS'p'
p54
aS'\xf3'
p55
aS't'
p56
aS'\xf7'
p57
aS'x'
p58
aS'\xfb'
p59
aS'|'
p60
aS'\xff'
p61
aS'\x80'
p62
aS'\x03'
p63
aS'\x84'
p64
aS'\x07'
p65
aS'\x88'
p66
aS'\x0b'
p67
aS'\x8c'
p68
aS'\x0f'
p69
aS'\x90'
p70
aS'\x13'
p71
aS'\x94'
p72
aS'\x17'
p73
aS'\x98'
p74
aS'\x1b'
p75
aS'\x9c'
p76
aS'\x1f'
p77
aS'\xa0'
p78
aS'#'
p79
aS'\xa4'
p80
aS"'"
p81
aS'\xa8'
p82
aS'+'
p83
aS'\xac'
p84
aS'/'
p85
aS'\xb0'
p86
aS'3'
p87
aS'\xb4'
p88
aS'7'
p89
aS'\xb8'
p90
aS';'
p91
aS'\xbc'
p92
aS'?'
p93
aS'\xc0'
p94
aS'C'
p95
aS'\xc4'
p96
aS'G'
p97
aS'\xc8'
p98
aS'K'
p99
aS'\xcc'
p100
aS'O'
p101
aS'\xd0'
p102
aS'S'
p103
aS'\xd4'
p104
aS'W'
p105
aS'\xd8'
p106
aS'['
p107
aS'\xdc'
p108
aS'_'
p109
aS'\xe0'
p110
aS'c'
p111
aS'\xe4'
p112
aS'g'
p113
aS'\xe8'
p114
aS'k'
p115
aS'\xec'
p116
aS'o'
p117
aS'\xf0'
p118
aS's'
p119
aS'\xf4'
p120
aS'w'
p121
aS'\xf8'
p122
aS'{'
p123
aS'\xfc'
p124
aS'\x7f'
p125
aS'\x81'
p126
aS'\x02'
p127
aS'\x85'
p128
aS'\x06'
p129
aS'\x89'
p130
aS'\n'
p131
aS'\x8d'
p132
aS'\x0e'
p133
aS'\x91'
p134
aS'\x12'
p135
aS'\x95'
p136
aS'\x16'
p137
aS'\x99'
p138
aS'\x1a'
p139
aS'\x9d'
p140
aS'\x1e'
p141
aS'\xa1'
p142
aS'"'
p143
aS'\xa5'
p144
aS'&'
p145
aS'\xa9'
p146
aS'*'
p147
aS'\xad'
p148
aS'.'
p149
aS'\xb1'
p150
aS'2'
p151
aS'\xb5'
p152
aS'6'
p153
aS'\xb9'
p154
aS':'
p155
aS'\xbd'
p156
aS'>'
p157
aS'\xc1'
p158
aS'B'
p159
aS'\xc5'
p160
aS'F'
p161
aS'\xc9'
p162
aS'J'
p163
aS'\xcd'
p164
aS'N'
p165
aS'\xd1'
p166
aS'R'
p167
aS'\xd5'
p168
aS'V'
p169
aS'\xd9'
p170
aS'Z'
p171
aS'\xdd'
p172
aS'^'
p173
aS'\xe1'
p174
aS'b'
p175
aS'\xe5'
p176
aS'f'
p177
aS'\xe9'
p178
aS'j'
p179
aS'\xed'
p180
aS'n'
p181
aS'\xf1'
p182
aS'r'
p183
aS'\xf5'
p184
aS'v'
p185
aS'\xf9'
p186
aS'z'
p187
aS'\xfd'
p188
aS'~'
p189
aS'\x01'
p190
aS'\x82'
p191
aS'\x05'
p192
aS'\x86'
p193
aS'\t'
p194
aS'\x8a'
p195
aS'\x8e'
p196
aS'\x11'
p197
aS'\x92'
p198
aS'\x15'
p199
aS'\x96'
p200
aS'\x19'
p201
aS'\x9a'
p202
aS'\x1d'
p203
aS'\x9e'
p204
aS'!'
p205
aS'\xa2'
p206
aS'%'
p207
aS'\xa6'
p208
aS')'
p209
aS'\xaa'
p210
aS'-'
p211
aS'\xae'
p212
aS'1'
p213
aS'\xb2'
p214
aS'5'
p215
aS'\xb6'
p216
aS'9'
p217
aS'\xba'
p218
aS'='
p219
aS'\xbe'
p220
aS'A'
p221
aS'\xc2'
p222
aS'E'
p223
aS'\xc6'
p224
aS'I'
p225
aS'\xca'
p226
aS'M'
p227
aS'\xce'
p228
aS'Q'
p229
aS'\xd2'
p230
aS'U'
p231
aS'\xd6'
p232
aS'Y'
p233
aS'\xda'
p234
aS']'
p235
aS'\xde'
p236
aS'a'
p237
aS'\xe2'
p238
aS'e'
p239
aS'\xe6'
p240
aS'i'
p241
aS'\xea'
p242
aS'm'
p243
aS'\xee'
p244
aS'q'
p245
aS'\xf2'
p246
aS'u'
p247
aS'\xf6'
p248
aS'y'
p249
aS'\xfa'
p250
aS'}'
p251
aS'\xfe'
p252
a.

26
dbms/scripts/gen_benchmark_data/train.py
View File

@ -1,26 +0,0 @@
import argparse
from model import Model
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('--n_iter', type=int, default=10000,
help='number of iterations')
parser.add_argument('--save_dir', type=str, default='save',
help='dir for saving weights')
parser.add_argument('--data_path', type=str,
help='path to train data')
parser.add_argument('--learning_rate', type=int, default=0.0001,
help='learning rate')
parser.add_argument('--batch_size', type=int, default=64,
help='batch size')
parser.add_argument('--restore_from', type=str,
help='path to train saved weights')
args = parser.parse_args()
if __name__ == '__main__':
if not args.data_path:
raise Exception('please specify path to train data with --data_path')
gen = Model(args.learning_rate)
gen.train(args.data_path, args.save_dir, args.n_iter, args.batch_size, args.restore_from)

150
dbms/scripts/linear-counting-threshold.py
View File

@ -1,150 +0,0 @@
#!/usr/bin/python3.4
# -*- coding: utf-8 -*-
import sys
import argparse
import tempfile
import random
import subprocess
import bisect
from copy import deepcopy
# Псевдослучайный генератор уникальных чисел.
# http://preshing.com/20121224/how-to-generate-a-sequence-of-unique-random-integers/
class UniqueRandomGenerator:
prime = 4294967291
def __init__(self, seed_base, seed_offset):
self.index = self.permutePQR(self.permutePQR(seed_base) + 0x682f0161)
self.intermediate_offset = self.permutePQR(self.permutePQR(seed_offset) + 0x46790905)
def next(self):
val = self.permutePQR((self.permutePQR(self.index) + self.intermediate_offset) ^ 0x5bf03635)
self.index = self.index + 1
return val
def permutePQR(self, x):
if x >=self.prime:
return x
else:
residue = (x * x) % self.prime
if x <= self.prime/2:
return residue
else:
return self.prime - residue
# Создать таблицу содержащую уникальные значения.
def generate_data_source(host, port, http_port, min_cardinality, max_cardinality, count):
chunk_size = round((max_cardinality - (min_cardinality + 1)) / float(count))
used_values = 0
cur_count = 0
next_size = 0
sup = 32768
n1 = random.randrange(0, sup)
n2 = random.randrange(0, sup)
urng = UniqueRandomGenerator(n1, n2)
is_first = True
with tempfile.TemporaryDirectory() as tmp_dir:
filename = tmp_dir + '/table.txt'
with open(filename, 'w+b') as file_handle:
while cur_count < count:
if is_first == True:
is_first = False
if min_cardinality != 0:
next_size = min_cardinality + 1
else:
next_size = chunk_size
else:
next_size += chunk_size
while used_values < next_size:
h = urng.next()
used_values = used_values + 1
out = str(h) + "\t" + str(cur_count) + "\n";
file_handle.write(bytes(out, 'UTF-8'));
cur_count = cur_count + 1
query = "DROP TABLE IF EXISTS data_source"
subprocess.check_output(["clickhouse-client", "--host", host, "--port", str(port), "--query", query])
query = "CREATE TABLE data_source(UserID UInt64, KeyID UInt64) ENGINE=TinyLog"
subprocess.check_output(["clickhouse-client", "--host", host, "--port", str(port), "--query", query])
cat = subprocess.Popen(("cat", filename), stdout=subprocess.PIPE)
subprocess.check_output(("POST", "http://{0}:{1}/?query=INSERT INTO data_source FORMAT TabSeparated".format(host, http_port)), stdin=cat.stdout)
cat.wait()
def perform_query(host, port):
query = "SELECT runningAccumulate(uniqExactState(UserID)) AS exact, "
query += "runningAccumulate(uniqCombinedRawState(UserID)) AS raw, "
query += "runningAccumulate(uniqCombinedLinearCountingState(UserID)) AS linear_counting, "
query += "runningAccumulate(uniqCombinedBiasCorrectedState(UserID)) AS bias_corrected "
query += "FROM data_source GROUP BY KeyID"
return subprocess.check_output(["clickhouse-client", "--host", host, "--port", port, "--query", query])
def parse_clickhouse_response(response):
parsed = []
lines = response.decode().split("\n")
for cur_line in lines:
rows = cur_line.split("\t")
if len(rows) == 4:
parsed.append([float(rows[0]), float(rows[1]), float(rows[2]), float(rows[3])])
return parsed
def accumulate_data(accumulated_data, data):
if not accumulated_data:
accumulated_data = deepcopy(data)
else:
for row1, row2 in zip(accumulated_data, data):
row1[1] += row2[1];
row1[2] += row2[2];
row1[3] += row2[3];
return accumulated_data
def dump_graphs(data, count):
with open("raw_graph.txt", "w+b") as fh1, open("linear_counting_graph.txt", "w+b") as fh2, open("bias_corrected_graph.txt", "w+b") as fh3:
expected_tab = []
bias_tab = []
for row in data:
exact = row[0]
raw = row[1] / count;
linear_counting = row[2] / count;
bias_corrected = row[3] / count;
outstr = "{0}\t{1}\n".format(exact, abs(raw - exact) / exact)
fh1.write(bytes(outstr, 'UTF-8'))
outstr = "{0}\t{1}\n".format(exact, abs(linear_counting - exact) / exact)
fh2.write(bytes(outstr, 'UTF-8'))
outstr = "{0}\t{1}\n".format(exact, abs(bias_corrected - exact) / exact)
fh3.write(bytes(outstr, 'UTF-8'))
def start():
parser = argparse.ArgumentParser(description = "Generate graphs that help to determine the linear counting threshold.")
parser.add_argument("-x", "--host", default="localhost", help="clickhouse host name");
parser.add_argument("-p", "--port", type=int, default=9000, help="clickhouse client TCP port");
parser.add_argument("-t", "--http_port", type=int, default=8123, help="clickhouse HTTP port");
parser.add_argument("-i", "--iterations", type=int, default=5000, help="number of iterations");
parser.add_argument("-m", "--min_cardinality", type=int, default=16384, help="minimal cardinality");
parser.add_argument("-M", "--max_cardinality", type=int, default=655360, help="maximal cardinality");
args = parser.parse_args()
accumulated_data = []
for i in range(0, args.iterations):
print(i + 1)
sys.stdout.flush()
generate_data_source(args.host, str(args.port), str(args.http_port), args.min_cardinality, args.max_cardinality, 1000)
response = perform_query(args.host, str(args.port))
data = parse_clickhouse_response(response)
accumulated_data = accumulate_data(accumulated_data, data)
dump_graphs(accumulated_data, args.iterations)
if __name__ == "__main__": start()

10
dbms/scripts/merge_algorithm/add_parts.sh
View File

@ -1,10 +0,0 @@
#!/usr/bin/env bash
for (( i = 0; i < 1000; i++ )); do
if (( RANDOM % 10 )); then
clickhouse-client --port=9007 --query="INSERT INTO mt (x) SELECT rand64() AS x FROM system.numbers LIMIT 100000"
else
clickhouse-client --port=9007 --query="INSERT INTO mt (x) SELECT rand64() AS x FROM system.numbers LIMIT 300000"
fi
done

76
dbms/scripts/merge_algorithm/drawer.py
View File

@ -1,76 +0,0 @@
from __future__ import print_function
import argparse
import matplotlib.pyplot as plt
import ast
TMP_FILE='tmp.tsv'
def parse_args():
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('-f', '--file', default='data.tsv')
cfg = parser.parse_args()
return cfg
def draw():
place = dict()
max_coord = 0
global_top = 0
for line in open(TMP_FILE):
numbers = line.split('\t')
if len(numbers) <= 2:
continue
name = numbers[-2]
if numbers[0] == '1':
dx = int(numbers[3])
max_coord += dx
place[name] = [1, max_coord, 1, dx]
max_coord += dx
plt.plot([max_coord - 2 * dx, max_coord], [1, 1])
for line in open(TMP_FILE):
numbers = line.split('\t')
if len(numbers) <= 2:
continue
name = numbers[-2]
if numbers[0] == '2':
list = ast.literal_eval(numbers[-1])
coord = [0,0,0,0]
for cur_name in list:
coord[0] = max(place[cur_name][0], coord[0])
coord[1] += place[cur_name][1] * place[cur_name][2]
coord[2] += place[cur_name][2]
coord[3] += place[cur_name][3]
coord[1] /= coord[2]
coord[0] += 1
global_top = max(global_top, coord[0])
place[name] = coord
for cur_name in list:
plt.plot([coord[1], place[cur_name][1]],[coord[0], place[cur_name][0]])
plt.plot([coord[1] - coord[3], coord[1] + coord[3]], [coord[0], coord[0]])
plt.plot([0], [global_top + 1])
plt.plot([0], [-1])
plt.show()
def convert(input_file):
print(input_file)
tmp_file = open(TMP_FILE, "w")
for line in open(input_file):
numbers = line.split('\t')
numbers2 = numbers[-2].split('_')
if numbers2[-2] == numbers2[-3]:
numbers2[-2] = str(int(numbers2[-2]) + 1)
numbers2[-3] = str(int(numbers2[-3]) + 1)
numbers[-2] = '_'.join(numbers2[1:])
print('\t'.join(numbers), end='', file=tmp_file)
else:
print(line, end='', file=tmp_file)
def main():
cfg = parse_args()
convert(cfg.file)
draw()
if __name__ == '__main__':
main()

61
dbms/scripts/merge_algorithm/stats.py
View File

@ -1,61 +0,0 @@
import time
import ast
from datetime import datetime
FILE='data.tsv'
def get_metrix():
data = []
time_to_merge = 0
count_of_parts = 0
max_count_of_parts = 0
parts_in_time = []
last_date = 0
for line in open(FILE):
fields = line.split('\t')
last_date = datetime.strptime(fields[2], '%Y-%m-%d %H:%M:%S')
break
for line in open(FILE):
fields = line.split('\t')
cur_date = datetime.strptime(fields[2], '%Y-%m-%d %H:%M:%S')
if fields[0] == '2':
time_to_merge += int(fields[4])
list = ast.literal_eval(fields[-1])
count_of_parts -= len(list) - 1
else:
count_of_parts += 1
if max_count_of_parts < count_of_parts:
max_count_of_parts = count_of_parts
parts_in_time.append([(cur_date-last_date).total_seconds(), count_of_parts])
last_date = cur_date
stats_parts_in_time = []
global_time = 0
average_parts = 0
for i in range(max_count_of_parts + 1):
stats_parts_in_time.append(0)
for elem in parts_in_time:
stats_parts_in_time[elem[1]] += elem[0]
global_time += elem[0]
average_parts += elem[0] * elem[1]
for i in range(max_count_of_parts):
stats_parts_in_time[i] /= global_time
average_parts /= global_time
return time_to_merge, max_count_of_parts, average_parts, stats_parts_in_time
def main():
time_to_merge, max_parts, average_parts, stats_parts = get_metrix()
print('time_to_merge=', time_to_merge)
print('max_parts=', max_parts)
print('average_parts=', average_parts)
print('stats_parts=', stats_parts)
if __name__ == '__main__':
main()

56
dbms/scripts/test_intHash32_for_linear_counting.py
View File

@ -1,56 +0,0 @@
#!/usr/bin/python3
import sys
import math
import statistics as stat
start = int(sys.argv[1])
end = int(sys.argv[2])
#Copied from dbms/src/Common/HashTable/Hash.h
def intHash32(key, salt = 0):
key ^= salt;
key = (~key) + (key << 18);
key = key ^ ((key >> 31) | (key << 33));
key = key * 21;
key = key ^ ((key >> 11) | (key << 53));
key = key + (key << 6);
key = key ^ ((key >> 22) | (key << 42));
return key & 0xffffffff
#Number of buckets for precision p = 12, m = 2^p
m = 4096
n = start
c = 0
m1 = {}
m2 = {}
l1 = []
l2 = []
while n <= end:
c += 1
h = intHash32(n)
#Extract left most 12 bits
x1 = (h >> 20) & 0xfff
m1[x1] = 1
z1 = m - len(m1)
#Linear counting formula
u1 = int(m * math.log(float(m) / float(z1)))
e1 = abs(100*float(u1 - c)/float(c))
l1.append(e1)
print("%d %d %d %f" % (n, c, u1, e1))
#Extract right most 12 bits
x2 = h & 0xfff
m2[x2] = 1
z2 = m - len(m2)
u2 = int(m * math.log(float(m) / float(z2)))
e2 = abs(100*float(u2 - c)/float(c))
l2.append(e2)
print("%d %d %d %f" % (n, c, u2, e2))
n += 1
print("Left 12 bits error: min=%f max=%f avg=%f median=%f median_low=%f median_high=%f" % (min(l1), max(l1), stat.mean(l1), stat.median(l1), stat.median_low(l1), stat.median_high(l1)))
print("Right 12 bits error: min=%f max=%f avg=%f median=%f median_low=%f median_high=%f" % (min(l2), max(l2), stat.mean(l2), stat.median(l2), stat.median_low(l2), stat.median_high(l2)))

11
dbms/scripts/test_uniq_functions.sh
View File

@ -1,11 +0,0 @@
#!/usr/bin/env bash
for ((p = 2; p <= 10; p++))
do
for ((i = 1; i <= 9; i++))
do
n=$(( 10**p * i ))
echo -n "$n "
clickhouse-client -q "select uniqHLL12(number), uniq(number), uniqCombined(number) from numbers($n);"
done
done

0
dbms/src/AggregateFunctions/AggregateFunctionsStatisticsSimple.cpp → dbms/src/AggregateFunctions/AggregateFunctionStatisticsSimple.cpp
View File

4
dbms/src/Columns/ColumnAggregateFunction.h
View File

@ -43,13 +43,13 @@ using Arenas = std::vector<ArenaPtr>;
* specifying which individual values should be destroyed and which ones should not.
* Clearly, this method would have a substantially non-zero price.
*/
class ColumnAggregateFunction final : public COWPtrHelper<IColumn, ColumnAggregateFunction>
class ColumnAggregateFunction final : public COWHelper<IColumn, ColumnAggregateFunction>
{
public:
using Container = PaddedPODArray<AggregateDataPtr>;
private:
friend class COWPtrHelper<IColumn, ColumnAggregateFunction>;
friend class COWHelper<IColumn, ColumnAggregateFunction>;
/// Memory pools. Aggregate states are allocated from them.
Arenas arenas;

6
dbms/src/Columns/ColumnArray.h
View File

@ -13,10 +13,10 @@ namespace DB
* In memory, it is represented as one column of a nested type, whose size is equal to the sum of the sizes of all arrays,
* and as an array of offsets in it, which allows you to get each element.
*/
class ColumnArray final : public COWPtrHelper<IColumn, ColumnArray>
class ColumnArray final : public COWHelper<IColumn, ColumnArray>
{
private:
friend class COWPtrHelper<IColumn, ColumnArray>;
friend class COWHelper<IColumn, ColumnArray>;
/** Create an array column with specified values and offsets. */
ColumnArray(MutableColumnPtr && nested_column, MutableColumnPtr && offsets_column);
@ -30,7 +30,7 @@ public:
/** Create immutable column using immutable arguments. This arguments may be shared with other columns.
* Use IColumn::mutate in order to make mutable column and mutate shared nested columns.
*/
using Base = COWPtrHelper<IColumn, ColumnArray>;
using Base = COWHelper<IColumn, ColumnArray>;
static Ptr create(const ColumnPtr & nested_column, const ColumnPtr & offsets_column)
{

4
dbms/src/Columns/ColumnConst.h
View File

@ -18,10 +18,10 @@ namespace ErrorCodes
/** ColumnConst contains another column with single element,
* but looks like a column with arbitrary amount of same elements.
*/
class ColumnConst final : public COWPtrHelper<IColumn, ColumnConst>
class ColumnConst final : public COWHelper<IColumn, ColumnConst>
{
private:
friend class COWPtrHelper<IColumn, ColumnConst>;
friend class COWHelper<IColumn, ColumnConst>;
WrappedPtr data;
size_t s;

4
dbms/src/Columns/ColumnDecimal.h
View File

@ -55,13 +55,13 @@ private:
/// A ColumnVector for Decimals
template <typename T>
class ColumnDecimal final : public COWPtrHelper<ColumnVectorHelper, ColumnDecimal<T>>
class ColumnDecimal final : public COWHelper<ColumnVectorHelper, ColumnDecimal<T>>
{
static_assert(IsDecimalNumber<T>);
private:
using Self = ColumnDecimal;
friend class COWPtrHelper<ColumnVectorHelper, Self>;
friend class COWHelper<ColumnVectorHelper, Self>;
public:
using Container = DecimalPaddedPODArray<T>;

4
dbms/src/Columns/ColumnFixedString.h
View File

@ -13,10 +13,10 @@ namespace DB
/** A column of values of "fixed-length string" type.
* If you insert a smaller string, it will be padded with zero bytes.
*/
class ColumnFixedString final : public COWPtrHelper<ColumnVectorHelper, ColumnFixedString>
class ColumnFixedString final : public COWHelper<ColumnVectorHelper, ColumnFixedString>
{
public:
friend class COWPtrHelper<ColumnVectorHelper, ColumnFixedString>;
friend class COWHelper<ColumnVectorHelper, ColumnFixedString>;
using Chars = PaddedPODArray<UInt8>;

4
dbms/src/Columns/ColumnFunction.h
View File

@ -15,10 +15,10 @@ namespace DB
/** A column containing a lambda expression.
* Behaves like a constant-column. Contains an expression, but not input or output data.
*/
class ColumnFunction final : public COWPtrHelper<IColumn, ColumnFunction>
class ColumnFunction final : public COWHelper<IColumn, ColumnFunction>
{
private:
friend class COWPtrHelper<IColumn, ColumnFunction>;
friend class COWHelper<IColumn, ColumnFunction>;
ColumnFunction(size_t size, FunctionBasePtr function, const ColumnsWithTypeAndName & columns_to_capture);

6
dbms/src/Columns/ColumnLowCardinality.h
View File

@ -14,9 +14,9 @@ namespace ErrorCodes
extern const int ILLEGAL_COLUMN;
}
class ColumnLowCardinality final : public COWPtrHelper<IColumn, ColumnLowCardinality>
class ColumnLowCardinality final : public COWHelper<IColumn, ColumnLowCardinality>
{
friend class COWPtrHelper<IColumn, ColumnLowCardinality>;
friend class COWHelper<IColumn, ColumnLowCardinality>;
ColumnLowCardinality(MutableColumnPtr && column_unique, MutableColumnPtr && indexes, bool is_shared = false);
ColumnLowCardinality(const ColumnLowCardinality & other) = default;
@ -25,7 +25,7 @@ public:
/** Create immutable column using immutable arguments. This arguments may be shared with other columns.
* Use IColumn::mutate in order to make mutable column and mutate shared nested columns.
*/
using Base = COWPtrHelper<IColumn, ColumnLowCardinality>;
using Base = COWHelper<IColumn, ColumnLowCardinality>;
static Ptr create(const ColumnPtr & column_unique_, const ColumnPtr & indexes_, bool is_shared = false)
{
return ColumnLowCardinality::create(column_unique_->assumeMutable(), indexes_->assumeMutable(), is_shared);

4
dbms/src/Columns/ColumnNothing.h
View File

@ -6,10 +6,10 @@
namespace DB
{
class ColumnNothing final : public COWPtrHelper<IColumnDummy, ColumnNothing>
class ColumnNothing final : public COWHelper<IColumnDummy, ColumnNothing>
{
private:
friend class COWPtrHelper<IColumnDummy, ColumnNothing>;
friend class COWHelper<IColumnDummy, ColumnNothing>;
ColumnNothing(size_t s_)
{

6
dbms/src/Columns/ColumnNullable.h
View File

@ -20,10 +20,10 @@ using ConstNullMapPtr = const NullMap *;
/// over a bitmap because columns are usually stored on disk as compressed
/// files. In this regard, using a bitmap instead of a byte map would
/// greatly complicate the implementation with little to no benefits.
class ColumnNullable final : public COWPtrHelper<IColumn, ColumnNullable>
class ColumnNullable final : public COWHelper<IColumn, ColumnNullable>
{
private:
friend class COWPtrHelper<IColumn, ColumnNullable>;
friend class COWHelper<IColumn, ColumnNullable>;
ColumnNullable(MutableColumnPtr && nested_column_, MutableColumnPtr && null_map_);
ColumnNullable(const ColumnNullable &) = default;
@ -32,7 +32,7 @@ public:
/** Create immutable column using immutable arguments. This arguments may be shared with other columns.
* Use IColumn::mutate in order to make mutable column and mutate shared nested columns.
*/
using Base = COWPtrHelper<IColumn, ColumnNullable>;
using Base = COWHelper<IColumn, ColumnNullable>;
static Ptr create(const ColumnPtr & nested_column_, const ColumnPtr & null_map_)
{
return ColumnNullable::create(nested_column_->assumeMutable(), null_map_->assumeMutable());

4
dbms/src/Columns/ColumnSet.h
View File

@ -14,10 +14,10 @@ using ConstSetPtr = std::shared_ptr<const Set>;
* Behaves like a constant-column (because the set is one, not its own for each line).
* This column has a nonstandard value, so it can not be obtained via a normal interface.
*/
class ColumnSet final : public COWPtrHelper<IColumnDummy, ColumnSet>
class ColumnSet final : public COWHelper<IColumnDummy, ColumnSet>
{
private:
friend class COWPtrHelper<IColumnDummy, ColumnSet>;
friend class COWHelper<IColumnDummy, ColumnSet>;
ColumnSet(size_t s_, const ConstSetPtr & data_) : data(data_) { s = s_; }
ColumnSet(const ColumnSet &) = default;

4
dbms/src/Columns/ColumnString.h
View File

@ -18,14 +18,14 @@ namespace DB
/** Column for String values.
*/
class ColumnString final : public COWPtrHelper<IColumn, ColumnString>
class ColumnString final : public COWHelper<IColumn, ColumnString>
{
public:
using Char = UInt8;
using Chars = PaddedPODArray<UInt8>;
private:
friend class COWPtrHelper<IColumn, ColumnString>;
friend class COWHelper<IColumn, ColumnString>;
/// Maps i'th position to offset to i+1'th element. Last offset maps to the end of all chars (is the size of all chars).
Offsets offsets;

6
dbms/src/Columns/ColumnTuple.h
View File

@ -12,10 +12,10 @@ namespace DB
* Mixed constant/non-constant columns is prohibited in tuple
* for implementation simplicity.
*/
class ColumnTuple final : public COWPtrHelper<IColumn, ColumnTuple>
class ColumnTuple final : public COWHelper<IColumn, ColumnTuple>
{
private:
friend class COWPtrHelper<IColumn, ColumnTuple>;
friend class COWHelper<IColumn, ColumnTuple>;
using TupleColumns = std::vector<WrappedPtr>;
TupleColumns columns;
@ -30,7 +30,7 @@ public:
/** Create immutable column using immutable arguments. This arguments may be shared with other columns.
* Use IColumn::mutate in order to make mutable column and mutate shared nested columns.
*/
using Base = COWPtrHelper<IColumn, ColumnTuple>;
using Base = COWHelper<IColumn, ColumnTuple>;
static Ptr create(const Columns & columns);
static Ptr create(const TupleColumns & columns);
static Ptr create(Columns && arg) { return create(arg); }

4
dbms/src/Columns/ColumnUnique.h
View File

@ -25,9 +25,9 @@ namespace ErrorCodes
}
template <typename ColumnType>
class ColumnUnique final : public COWPtrHelper<IColumnUnique, ColumnUnique<ColumnType>>
class ColumnUnique final : public COWHelper<IColumnUnique, ColumnUnique<ColumnType>>
{
friend class COWPtrHelper<IColumnUnique, ColumnUnique<ColumnType>>;
friend class COWHelper<IColumnUnique, ColumnUnique<ColumnType>>;
private:
explicit ColumnUnique(MutableColumnPtr && holder, bool is_nullable);

4
dbms/src/Columns/ColumnVector.h
View File

@ -90,13 +90,13 @@ template <> struct CompareHelper<Float64> : public FloatCompareHelper<Float64> {
/** A template for columns that use a simple array to store.
*/
template <typename T>
class ColumnVector final : public COWPtrHelper<ColumnVectorHelper, ColumnVector<T>>
class ColumnVector final : public COWHelper<ColumnVectorHelper, ColumnVector<T>>
{
static_assert(!IsDecimalNumber<T>);
private:
using Self = ColumnVector;
friend class COWPtrHelper<ColumnVectorHelper, Self>;
friend class COWHelper<ColumnVectorHelper, Self>;
struct less;
struct greater;

8
dbms/src/Columns/IColumn.h
View File

@ -1,7 +1,7 @@
#pragma once
#include <Core/Field.h>
#include <Common/COWPtr.h>
#include <Common/COW.h>
#include <Common/PODArray.h>
#include <Common/Exception.h>
#include <common/StringRef.h>
@ -24,13 +24,13 @@ class Arena;
class ColumnGathererStream;
/// Declares interface to store columns in memory.
class IColumn : public COWPtr<IColumn>
class IColumn : public COW<IColumn>
{
private:
friend class COWPtr<IColumn>;
friend class COW<IColumn>;
/// Creates the same column with the same data.
/// This is internal method to use from COWPtr.
/// This is internal method to use from COW.
/// It performs shallow copy with copy-ctor and not useful from outside.
/// If you want to copy column for modification, look at 'mutate' method.
virtual MutablePtr clone() const = 0;

38
dbms/src/Common/COWPtr.h → dbms/src/Common/COW.h
View File

@ -10,10 +10,10 @@
*
* Usage:
class Column : public COWPtr<Column>
class Column : public COW<Column>
{
private:
friend class COWPtr<Column>;
friend class COW<Column>;
/// Leave all constructors in private section. They will be avaliable through 'create' method.
Column();
@ -23,7 +23,7 @@
public:
/// Correctly use const qualifiers in your interface.
virtual ~IColumn() {}
virtual ~Column() {}
};
* It will provide 'create' and 'mutate' methods.
@ -63,7 +63,7 @@
* Actually it is, if your values are small or if copying is done implicitly.
* This is the case for string implementations.
*
* In contrast, COWPtr is intended for the cases when you need to share states of large objects,
* In contrast, COW is intended for the cases when you need to share states of large objects,
* (when you usually will use std::shared_ptr) but you also want precise control over modification
* of this shared state.
*
@ -73,7 +73,7 @@
* to use std::unique_ptr for it somehow.
*/
template <typename Derived>
class COWPtr : public boost::intrusive_ref_counter<Derived>
class COW : public boost::intrusive_ref_counter<Derived>
{
private:
Derived * derived() { return static_cast<Derived *>(this); }
@ -96,8 +96,8 @@ protected:
private:
using Base = IntrusivePtr<T>;
template <typename> friend class COWPtr;
template <typename, typename> friend class COWPtrHelper;
template <typename> friend class COW;
template <typename, typename> friend class COWHelper;
explicit mutable_ptr(T * ptr) : Base(ptr) {}
@ -115,7 +115,7 @@ protected:
mutable_ptr() = default;
mutable_ptr(const std::nullptr_t *) {}
mutable_ptr(std::nullptr_t) {}
};
public:
@ -128,8 +128,8 @@ protected:
private:
using Base = IntrusivePtr<const T>;
template <typename> friend class COWPtr;
template <typename, typename> friend class COWPtrHelper;
template <typename> friend class COW;
template <typename, typename> friend class COWHelper;
explicit immutable_ptr(const T * ptr) : Base(ptr) {}
@ -159,7 +159,7 @@ protected:
immutable_ptr() = default;
immutable_ptr(const std::nullptr_t *) {}
immutable_ptr(std::nullptr_t) {}
};
public:
@ -192,7 +192,7 @@ public:
MutablePtr assumeMutable() const
{
return const_cast<COWPtr*>(this)->getPtr();
return const_cast<COW*>(this)->getPtr();
}
Derived & assumeMutableRef() const
@ -244,7 +244,7 @@ public:
*
* NOTE:
* If you override 'mutate' method in inherited classes, don't forget to make it virtual in base class or to make it call a virtual method.
* (COWPtr itself doesn't force any methods to be virtual).
* (COW itself doesn't force any methods to be virtual).
*
* See example in "cow_compositions.cpp".
*/
@ -255,22 +255,22 @@ public:
/** Helper class to support inheritance.
* Example:
*
* class IColumn : public COWPtr<IColumn>
* class IColumn : public COW<IColumn>
* {
* friend class COWPtr<IColumn>;
* friend class COW<IColumn>;
* virtual MutablePtr clone() const = 0;
* virtual ~IColumn() {}
* };
*
* class ConcreteColumn : public COWPtrHelper<IColumn, ConcreteColumn>
* class ConcreteColumn : public COWHelper<IColumn, ConcreteColumn>
* {
* friend class COWPtrHelper<IColumn, ConcreteColumn>;
* friend class COWHelper<IColumn, ConcreteColumn>;
* };
*
* Here is complete inheritance diagram:
*
* ConcreteColumn
* COWPtrHelper<IColumn, ConcreteColumn>
* COWHelper<IColumn, ConcreteColumn>
* IColumn
* CowPtr<IColumn>
* boost::intrusive_ref_counter<IColumn>
@ -278,7 +278,7 @@ public:
* See example in "cow_columns.cpp".
*/
template <typename Base, typename Derived>
class COWPtrHelper : public Base
class COWHelper : public Base
{
private:
Derived * derived() { return static_cast<Derived *>(this); }

2
dbms/src/Common/HashTable/HashTable.h
View File

@ -566,7 +566,7 @@ public:
DB::ReadBuffer & in;
Cell cell;
size_t read_count = 0;
size_t size;
size_t size = 0;
bool is_eof = false;
bool is_initialized = false;
};

2
dbms/src/Common/ZooKeeper/ZooKeeperNodeCache.h
View File

@ -41,7 +41,7 @@ public:
{
bool exists = false;
std::string contents;
Coordination::Stat stat;
Coordination::Stat stat{};
};
ZNode get(const std::string & path, EventPtr watch_event);

10
dbms/src/Common/tests/cow_columns.cpp
View File

@ -1,11 +1,11 @@
#include <Common/COWPtr.h>
#include <Common/COW.h>
#include <iostream>
class IColumn : public COWPtr<IColumn>
class IColumn : public COW<IColumn>
{
private:
friend class COWPtr<IColumn>;
friend class COW<IColumn>;
virtual MutablePtr clone() const = 0;
public:
@ -22,10 +22,10 @@ public:
using ColumnPtr = IColumn::Ptr;
using MutableColumnPtr = IColumn::MutablePtr;
class ConcreteColumn : public COWPtrHelper<IColumn, ConcreteColumn>
class ConcreteColumn : public COWHelper<IColumn, ConcreteColumn>
{
private:
friend class COWPtrHelper<IColumn, ConcreteColumn>;
friend class COWHelper<IColumn, ConcreteColumn>;
int data;
ConcreteColumn(int data) : data(data) {}

14
dbms/src/Common/tests/cow_compositions.cpp
View File

@ -1,11 +1,11 @@
#include <Common/COWPtr.h>
#include <Common/COW.h>
#include <iostream>
class IColumn : public COWPtr<IColumn>
class IColumn : public COW<IColumn>
{
private:
friend class COWPtr<IColumn>;
friend class COW<IColumn>;
virtual MutablePtr clone() const = 0;
virtual MutablePtr deepMutate() const { return shallowMutate(); }
@ -24,10 +24,10 @@ public:
using ColumnPtr = IColumn::Ptr;
using MutableColumnPtr = IColumn::MutablePtr;
class ConcreteColumn : public COWPtrHelper<IColumn, ConcreteColumn>
class ConcreteColumn : public COWHelper<IColumn, ConcreteColumn>
{
private:
friend class COWPtrHelper<IColumn, ConcreteColumn>;
friend class COWHelper<IColumn, ConcreteColumn>;
int data;
ConcreteColumn(int data) : data(data) {}
@ -38,10 +38,10 @@ public:
void set(int value) override { data = value; }
};
class ColumnComposition : public COWPtrHelper<IColumn, ColumnComposition>
class ColumnComposition : public COWHelper<IColumn, ColumnComposition>
{
private:
friend class COWPtrHelper<IColumn, ColumnComposition>;
friend class COWHelper<IColumn, ColumnComposition>;
ConcreteColumn::WrappedPtr wrapped;

2
dbms/src/Core/SortCursor.h
View File

@ -29,7 +29,7 @@ struct SortCursorImpl
* Cursor number (always?) equals to number of merging part.
* Therefore this field can be used to determine part number of current row (see ColumnGathererStream).
*/
size_t order;
size_t order = 0;
using NeedCollationFlags = std::vector<UInt8>;

2
dbms/src/Core/iostream_debug_helpers.cpp
View File

@ -12,7 +12,7 @@
#include <Interpreters/ExpressionAnalyzer.h>
#include <Parsers/IAST.h>
#include <Storages/IStorage.h>
#include <Common/COWPtr.h>
#include <Common/COW.h>
#include <Common/FieldVisitors.h>
namespace DB

2
dbms/src/DataStreams/CollapsingFinalBlockInputStream.h
View File

@ -163,7 +163,7 @@ private:
struct Cursor
{
MergingBlockPtr block;
size_t pos;
size_t pos = 0;
Cursor() {}
explicit Cursor(const MergingBlockPtr & block_, size_t pos_ = 0) : block(block_), pos(pos_) {}

2
dbms/src/DataStreams/MergingSortedBlockInputStream.h
View File

@ -81,7 +81,7 @@ protected:
struct RowRef
{
ColumnRawPtrs * columns = nullptr;
size_t row_num;
size_t row_num = 0;
SharedBlockPtr shared_block;
void swap(RowRef & other)

2
dbms/src/DataStreams/ParallelInputsProcessor.h
View File

@ -163,7 +163,7 @@ private:
struct InputData
{
BlockInputStreamPtr in;
size_t i; /// The source number (for debugging).
size_t i = 0; /// The source number (for debugging).
InputData() {}
InputData(const BlockInputStreamPtr & in_, size_t i_) : in(in_), i(i_) {}

6
dbms/src/DataTypes/IDataType.h
View File

@ -1,7 +1,7 @@
#pragma once
#include <memory>
#include <Common/COWPtr.h>
#include <Common/COW.h>
#include <boost/noncopyable.hpp>
#include <Core/Field.h>
@ -17,8 +17,8 @@ class IDataType;
struct FormatSettings;
class IColumn;
using ColumnPtr = COWPtr<IColumn>::Ptr;
using MutableColumnPtr = COWPtr<IColumn>::MutablePtr;
using ColumnPtr = COW<IColumn>::Ptr;
using MutableColumnPtr = COW<IColumn>::MutablePtr;
using DataTypePtr = std::shared_ptr<const IDataType>;
using DataTypes = std::vector<DataTypePtr>;

2
dbms/src/Dictionaries/ExternalQueryBuilder.cpp
View File

@ -298,7 +298,7 @@ ExternalQueryBuilder::composeLoadKeysQuery(const Columns & key_columns, const st
composeKeyCondition(key_columns, row, out);
}
}
else if (method == IN_WITH_TUPLES)
else /* if (method == IN_WITH_TUPLES) */
{
writeString(composeKeyTupleDefinition(), out);
writeString(" IN (", out);

2
dbms/src/IO/HexWriteBuffer.h
View File

@ -15,7 +15,7 @@ namespace DB
class HexWriteBuffer final : public WriteBuffer
{
protected:
char buf[DBMS_HEX_WRITE_BUFFER_SIZE];
char buf[DBMS_HEX_WRITE_BUFFER_SIZE]; //-V730
WriteBuffer & out;
void nextImpl() override;

2
dbms/src/IO/readFloatText.h
View File

@ -312,7 +312,7 @@ template <typename T, typename ReturnType>
ReturnType readFloatTextFastImpl(T & x, ReadBuffer & in)
{
static_assert(std::is_same_v<T, double> || std::is_same_v<T, float>, "Argument for readFloatTextImpl must be float or double");
static_assert('a' > '.' && 'A' > '.' && '\n' < '.' && '\t' < '.' && '\'' < '.' && '"' < '.', "Layout of char is not like ASCII");
static_assert('a' > '.' && 'A' > '.' && '\n' < '.' && '\t' < '.' && '\'' < '.' && '"' < '.', "Layout of char is not like ASCII"); //-V501
static constexpr bool throw_exception = std::is_same_v<ReturnType, void>;

4
dbms/src/Interpreters/Aggregator.h
View File

@ -407,12 +407,12 @@ struct AggregatedDataVariants : private boost::noncopyable
*/
Aggregator * aggregator = nullptr;
size_t keys_size; /// Number of keys. NOTE do we need this field?
size_t keys_size{}; /// Number of keys. NOTE do we need this field?
Sizes key_sizes; /// Dimensions of keys, if keys of fixed length
/// Pools for states of aggregate functions. Ownership will be later transferred to ColumnAggregateFunction.
Arenas aggregates_pools;
Arena * aggregates_pool; /// The pool that is currently used for allocation.
Arena * aggregates_pool{}; /// The pool that is currently used for allocation.
/** Specialization for the case when there are no keys, and for keys not fitted into max_rows_to_group_by.
*/

4
dbms/src/Interpreters/Cluster.h
View File

@ -112,8 +112,8 @@ public:
/// Name of directory for asynchronous write to StorageDistributed if has_internal_replication
std::string dir_name_for_internal_replication;
/// Number of the shard, the indexation begins with 1
UInt32 shard_num;
UInt32 weight;
UInt32 shard_num = 0;
UInt32 weight = 1;
Addresses local_addresses;
/// nullptr if there are no remote addresses
ConnectionPoolWithFailoverPtr pool;

1
dbms/src/Interpreters/ExpressionAnalyzer.cpp
View File

@ -36,7 +36,6 @@
#include <AggregateFunctions/parseAggregateFunctionParameters.h>
#include <Storages/StorageDistributed.h>
#include <Storages/StorageMemory.h>
#include <Storages/StorageJoin.h>
#include <DataStreams/copyData.h>

2
dbms/src/Interpreters/ExpressionJIT.cpp
View File

@ -77,7 +77,7 @@ namespace
{
const char * data = nullptr;
const char * null = nullptr;
size_t stride;
size_t stride = 0;
};
struct ColumnDataPlaceholder

31
dbms/src/Storages/Kafka/KafkaBlockInputStream.cpp
View File

@ -24,36 +24,21 @@ KafkaBlockInputStream::~KafkaBlockInputStream()
return;
if (broken)
{
LOG_TRACE(storage.log, "Re-joining claimed consumer after failure");
consumer->unsubscribe();
}
buffer->subBufferAs<ReadBufferFromKafkaConsumer>()->unsubscribe();
storage.pushConsumer(consumer);
storage.pushBuffer(buffer);
}
void KafkaBlockInputStream::readPrefixImpl()
{
consumer = storage.tryClaimConsumer(context.getSettingsRef().queue_max_wait_ms.totalMilliseconds());
claimed = !!consumer;
buffer = storage.tryClaimBuffer(context.getSettingsRef().queue_max_wait_ms.totalMilliseconds());
claimed = !!buffer;
if (!consumer)
consumer = std::make_shared<cppkafka::Consumer>(storage.createConsumerConfiguration());
if (!buffer)
buffer = storage.createBuffer();
// While we wait for an assignment after subscribtion, we'll poll zero messages anyway.
// If we're doing a manual select then it's better to get something after a wait, then immediate nothing.
if (consumer->get_subscription().empty())
{
using namespace std::chrono_literals;
buffer->subBufferAs<ReadBufferFromKafkaConsumer>()->subscribe(storage.topics);
consumer->pause(); // don't accidentally read any messages
consumer->subscribe(storage.topics);
consumer->poll(5s);
consumer->resume();
}
buffer = std::make_unique<DelimitedReadBuffer>(
new ReadBufferFromKafkaConsumer(consumer, storage.log, max_block_size), storage.row_delimiter);
addChild(FormatFactory::instance().getInput(storage.format_name, *buffer, storage.getSampleBlock(), context, max_block_size));
broken = true;
@ -66,4 +51,4 @@ void KafkaBlockInputStream::readSuffixImpl()
broken = false;
}
} // namespace DB
}

4
dbms/src/Storages/Kafka/KafkaBlockInputStream.h
View File

@ -1,7 +1,6 @@
#pragma once
#include <DataStreams/IBlockInputStream.h>
#include <IO/DelimitedReadBuffer.h>
#include <Interpreters/Context.h>
#include <Storages/Kafka/StorageKafka.h>
@ -27,8 +26,7 @@ private:
Context context;
UInt64 max_block_size;
ConsumerPtr consumer;
std::unique_ptr<DelimitedReadBuffer> buffer;
BufferPtr buffer;
bool broken = true, claimed = false;
};

27
dbms/src/Storages/Kafka/ReadBufferFromKafkaConsumer.cpp
View File

@ -2,10 +2,11 @@
namespace DB
{
namespace
{
const auto READ_POLL_MS = 500; /// How long to wait for a batch of messages.
} // namespace
}
void ReadBufferFromKafkaConsumer::commit()
{
@ -13,10 +14,32 @@ void ReadBufferFromKafkaConsumer::commit()
return;
auto & previous = *std::prev(current);
LOG_TRACE(log, "Committing message with offset " << previous.get_offset());
consumer->async_commit(previous);
}
void ReadBufferFromKafkaConsumer::subscribe(const Names & topics)
{
// While we wait for an assignment after subscribtion, we'll poll zero messages anyway.
// If we're doing a manual select then it's better to get something after a wait, then immediate nothing.
if (consumer->get_subscription().empty())
{
using namespace std::chrono_literals;
consumer->pause(); // don't accidentally read any messages
consumer->subscribe(topics);
consumer->poll(5s);
consumer->resume();
}
}
void ReadBufferFromKafkaConsumer::unsubscribe()
{
LOG_TRACE(log, "Re-joining claimed consumer after failure");
consumer->unsubscribe();
}
/// Do commit messages implicitly after we processed the previous batch.
bool ReadBufferFromKafkaConsumer::nextImpl()
{
@ -50,4 +73,4 @@ bool ReadBufferFromKafkaConsumer::nextImpl()
return true;
}
} // namespace DB
}

13
dbms/src/Storages/Kafka/ReadBufferFromKafkaConsumer.h
View File

@ -1,14 +1,18 @@
#pragma once
#include <IO/ReadBuffer.h>
#include <Core/Names.h>
#include <IO/DelimitedReadBuffer.h>
#include <common/logger_useful.h>
#include <cppkafka/cppkafka.h>
namespace DB
{
using BufferPtr = std::shared_ptr<DelimitedReadBuffer>;
using ConsumerPtr = std::shared_ptr<cppkafka::Consumer>;
class ReadBufferFromKafkaConsumer : public ReadBuffer
{
public:
@ -17,8 +21,9 @@ public:
{
}
// Commit all processed messages.
void commit();
void commit(); // Commit all processed messages.
void subscribe(const Names & topics); // Subscribe internal consumer to topics.
void unsubscribe(); // Unsubscribe internal consumer in case of failure.
private:
using Messages = std::vector<cppkafka::Message>;
@ -33,4 +38,4 @@ private:
bool nextImpl() override;
};
} // namespace DB
}

50
dbms/src/Storages/Kafka/StorageKafka.cpp
View File

@ -81,7 +81,7 @@ StorageKafka::StorageKafka(
row_delimiter(row_delimiter_),
schema_name(global_context.getMacros()->expand(schema_name_)),
num_consumers(num_consumers_), max_block_size(max_block_size_), log(&Logger::get("StorageKafka (" + table_name_ + ")")),
semaphore(0, num_consumers_), mutex(), consumers(),
semaphore(0, num_consumers_),
skip_broken(skip_broken_)
{
task = global_context.getSchedulePool().createTask(log->name(), [this]{ streamThread(); });
@ -124,12 +124,8 @@ void StorageKafka::startup()
{
for (size_t i = 0; i < num_consumers; ++i)
{
// Create a consumer and subscribe to topics
auto consumer = std::make_shared<cppkafka::Consumer>(createConsumerConfiguration());
consumer->subscribe(topics);
// Make consumer available
pushConsumer(consumer);
// Make buffer available
pushBuffer(createBuffer());
++num_created_consumers;
}
@ -146,8 +142,8 @@ void StorageKafka::shutdown()
// Close all consumers
for (size_t i = 0; i < num_created_consumers; ++i)
{
auto consumer = claimConsumer();
// FIXME: not sure if really close consumers here, and if we really need to close them here.
auto buffer = claimBuffer();
// FIXME: not sure if we really close consumers here, and if we really need to close them here.
}
LOG_TRACE(log, "Waiting for cleanup");
@ -203,14 +199,29 @@ cppkafka::Configuration StorageKafka::createConsumerConfiguration()
return conf;
}
ConsumerPtr StorageKafka::claimConsumer()
BufferPtr StorageKafka::createBuffer()
{
return tryClaimConsumer(-1L);
// Create a consumer and subscribe to topics
auto consumer = std::make_shared<cppkafka::Consumer>(createConsumerConfiguration());
consumer->subscribe(topics);
// Limit the number of batched messages to allow early cancellations
const Settings & settings = global_context.getSettingsRef();
size_t batch_size = max_block_size;
if (!batch_size)
batch_size = settings.max_block_size.value;
return std::make_shared<DelimitedReadBuffer>(new ReadBufferFromKafkaConsumer(consumer, log, batch_size), row_delimiter);
}
ConsumerPtr StorageKafka::tryClaimConsumer(long wait_ms)
BufferPtr StorageKafka::claimBuffer()
{
// Wait for the first free consumer
return tryClaimBuffer(-1L);
}
BufferPtr StorageKafka::tryClaimBuffer(long wait_ms)
{
// Wait for the first free buffer
if (wait_ms >= 0)
{
if (!semaphore.tryWait(wait_ms))
@ -219,17 +230,17 @@ ConsumerPtr StorageKafka::tryClaimConsumer(long wait_ms)
else
semaphore.wait();
// Take the first available consumer from the list
// Take the first available buffer from the list
std::lock_guard lock(mutex);
auto consumer = consumers.back();
consumers.pop_back();
return consumer;
auto buffer = buffers.back();
buffers.pop_back();
return buffer;
}
void StorageKafka::pushConsumer(ConsumerPtr consumer)
void StorageKafka::pushBuffer(BufferPtr buffer)
{
std::lock_guard lock(mutex);
consumers.push_back(consumer);
buffers.push_back(buffer);
semaphore.set();
}
@ -303,7 +314,6 @@ bool StorageKafka::streamToViews()
insert->table = table_name;
insert->no_destination = true; // Only insert into dependent views
// Limit the number of batched messages to allow early cancellations
const Settings & settings = global_context.getSettingsRef();
size_t block_size = max_block_size;
if (block_size == 0)

18
dbms/src/Storages/Kafka/StorageKafka.h
View File

@ -4,6 +4,7 @@
#include <Core/NamesAndTypes.h>
#include <DataStreams/IBlockOutputStream.h>
#include <Storages/IStorage.h>
#include <Storages/Kafka/ReadBufferFromKafkaConsumer.h>
#include <Poco/Event.h>
#include <Poco/Semaphore.h>
#include <ext/shared_ptr_helper.h>
@ -14,15 +15,13 @@
namespace DB
{
using ConsumerPtr = std::shared_ptr<cppkafka::Consumer>;
/** Implements a Kafka queue table engine that can be used as a persistent queue / buffer,
* or as a basic building block for creating pipelines with a continuous insertion / ETL.
*/
class StorageKafka : public ext::shared_ptr_helper<StorageKafka>, public IStorage
{
friend class KafkaBlockInputStream;
friend class KafkaBlockOutputStream;
friend class KafkaBlockInputStream;
friend class KafkaBlockOutputStream;
public:
std::string getName() const override { return "Kafka"; }
@ -40,7 +39,7 @@ public:
size_t max_block_size,
unsigned num_streams) override;
void rename(const String & /*new_path_to_db*/, const String & new_database_name, const String & new_table_name) override
void rename(const String & /* new_path_to_db */, const String & new_database_name, const String & new_table_name) override
{
table_name = new_table_name;
database_name = new_database_name;
@ -74,7 +73,7 @@ private:
// Consumer list
Poco::Semaphore semaphore;
std::mutex mutex;
std::vector<ConsumerPtr> consumers; /// Available consumers
std::vector<BufferPtr> buffers; /// available buffers for Kafka consumers
size_t skip_broken;
@ -83,9 +82,10 @@ private:
std::atomic<bool> stream_cancelled{false};
cppkafka::Configuration createConsumerConfiguration();
ConsumerPtr claimConsumer();
ConsumerPtr tryClaimConsumer(long wait_ms);
void pushConsumer(ConsumerPtr c);
BufferPtr createBuffer();
BufferPtr claimBuffer();
BufferPtr tryClaimBuffer(long wait_ms);
void pushBuffer(BufferPtr buf);
void streamThread();
bool streamToViews();

4
dbms/src/Storages/MergeTree/BoolMask.h
View File

@ -3,8 +3,8 @@
/// Multiple Boolean values. That is, two Boolean values: can it be true, can it be false.
struct BoolMask
{
bool can_be_true;
bool can_be_false;
bool can_be_true = false;
bool can_be_false = false;
BoolMask() {}
BoolMask(bool can_be_true_, bool can_be_false_) : can_be_true(can_be_true_), can_be_false(can_be_false_) {}

330
dbms/tests/clickhouse-test
View File

@ -20,6 +20,8 @@ from errno import ESRCH
import termcolor
from random import random
import commands
from multiprocessing import Pool
from contextlib import closing
MESSAGES_TO_RETRY = [
@ -87,34 +89,7 @@ def get_server_pid(server_tcp_port):
except Exception as ex:
return None
def main(args):
SERVER_DIED = False
def colored(text, color=None, on_color=None, attrs=None):
if sys.stdout.isatty() or args.force_color:
return termcolor.colored(text, color, on_color, attrs)
else:
return text
OP_SQUARE_BRACKET = colored("[", attrs=['bold'])
CL_SQUARE_BRACKET = colored("]", attrs=['bold'])
MSG_FAIL = OP_SQUARE_BRACKET + colored(" FAIL ", "red", attrs=['bold']) + CL_SQUARE_BRACKET
MSG_UNKNOWN = OP_SQUARE_BRACKET + colored(" UNKNOWN ", "yellow", attrs=['bold']) + CL_SQUARE_BRACKET
MSG_OK = OP_SQUARE_BRACKET + colored(" OK ", "green", attrs=['bold']) + CL_SQUARE_BRACKET
MSG_SKIPPED = OP_SQUARE_BRACKET + colored(" SKIPPED ", "cyan", attrs=['bold']) + CL_SQUARE_BRACKET
def is_data_present():
clickhouse_proc = Popen(shlex.split(args.client), stdin=PIPE, stdout=PIPE, stderr=PIPE)
(stdout, stderr) = clickhouse_proc.communicate("EXISTS TABLE test.hits")
if clickhouse_proc.returncode != 0:
raise CalledProcessError(clickhouse_proc.returncode, args.client, stderr)
return stdout.startswith('1')
def dump_report(destination, suite, test_case, report):
def dump_report(destination, suite, test_case, report):
if destination is not None:
destination_file = os.path.join(destination, suite, test_case + ".xml")
destination_dir = os.path.dirname(destination_file)
@ -127,129 +102,39 @@ def main(args):
report_root.append(report_suite)
report_file.write(et.tostring(report_root, encoding = "UTF-8", xml_declaration=True, pretty_print=True))
base_dir = os.path.abspath(args.queries)
tmp_dir = os.path.abspath(args.tmp)
# Keep same default values as in queries/shell_config.sh
os.environ.setdefault("CLICKHOUSE_BINARY", args.binary)
#os.environ.setdefault("CLICKHOUSE_CLIENT", args.client)
os.environ.setdefault("CLICKHOUSE_CONFIG", args.configserver)
if args.configclient:
os.environ.setdefault("CLICKHOUSE_CONFIG_CLIENT", args.configclient)
os.environ.setdefault("CLICKHOUSE_TMP", tmp_dir)
os.environ.setdefault("CLICKHOUSE_DATABASE", args.database)
# Force to print server warnings in stderr
# Shell scripts could change logging level
server_logs_level = "warning"
os.environ.setdefault("CLICKHOUSE_CLIENT_SERVER_LOGS_LEVEL", server_logs_level)
if args.zookeeper is None:
code, out = commands.getstatusoutput(args.extract_from_config +" --try --config " + args.configserver + ' --key zookeeper | grep . | wc -l')
try:
if int(out) > 0:
args.zookeeper = True
def colored(text, args, color=None, on_color=None, attrs=None):
if sys.stdout.isatty() or args.force_color:
return termcolor.colored(text, color, on_color, attrs)
else:
args.zookeeper = False
except ValueError:
args.zookeeper = False
return text
if args.shard is None:
code, out = commands.getstatusoutput(args.extract_from_config + " --try --config " + args.configserver + ' --key listen_host | grep -E "127.0.0.2|::"')
if out:
args.shard = True
else:
args.shard = False
SERVER_DIED = False
exit_code = 0
#def run_tests_array(all_tests, suite, suite_dir, suite_tmp_dir, run_total):
def run_tests_array(all_tests_with_params):
all_tests, suite, suite_dir, suite_tmp_dir, run_total = all_tests_with_params
global exit_code
global SERVER_DIED
OP_SQUARE_BRACKET = colored("[", args, attrs=['bold'])
CL_SQUARE_BRACKET = colored("]", args, attrs=['bold'])
MSG_FAIL = OP_SQUARE_BRACKET + colored(" FAIL ", args, "red", attrs=['bold']) + CL_SQUARE_BRACKET
MSG_UNKNOWN = OP_SQUARE_BRACKET + colored(" UNKNOWN ", args, "yellow", attrs=['bold']) + CL_SQUARE_BRACKET
MSG_OK = OP_SQUARE_BRACKET + colored(" OK ", args, "green", attrs=['bold']) + CL_SQUARE_BRACKET
MSG_SKIPPED = OP_SQUARE_BRACKET + colored(" SKIPPED ", args, "cyan", attrs=['bold']) + CL_SQUARE_BRACKET
passed_total = 0
skipped_total = 0
failures_total = 0
clickhouse_proc_create = Popen(shlex.split(args.client), stdin=PIPE, stdout=PIPE, stderr=PIPE)
clickhouse_proc_create.communicate("CREATE DATABASE IF NOT EXISTS " + args.database)
if args.database != "test":
clickhouse_proc_create = Popen(shlex.split(args.client), stdin=PIPE, stdout=PIPE, stderr=PIPE)
clickhouse_proc_create.communicate("CREATE DATABASE IF NOT EXISTS test")
def is_test_from_dir(suite_dir, case):
case_file = os.path.join(suite_dir, case)
(name, ext) = os.path.splitext(case)
return os.path.isfile(case_file) and (ext == '.sql' or ext == '.sh' or ext == '.py')
def sute_key_func(item):
if args.order == 'random':
return random()
if -1 == item.find('_'):
return 99998
prefix, suffix = item.split('_', 1)
try:
return int(prefix), suffix
except ValueError:
return 99997
for suite in sorted(os.listdir(base_dir), key=sute_key_func):
if SERVER_DIED:
break
suite_dir = os.path.join(base_dir, suite)
suite_re_obj = re.search('^[0-9]+_(.*)$', suite)
if not suite_re_obj: #skip .gitignore and so on
continue
suite_tmp_dir = os.path.join(tmp_dir, suite)
if not os.path.exists(suite_tmp_dir):
os.makedirs(suite_tmp_dir)
suite = suite_re_obj.group(1)
if os.path.isdir(suite_dir):
failures = 0
failures_chain = 0
if 'stateful' in suite and not is_data_present():
print("Won't run stateful tests because test data wasn't loaded.")
continue
if 'stateless' in suite and args.no_stateless:
print("Won't run stateless tests because they were manually disabled.")
continue
if 'stateful' in suite and args.no_stateful:
print("Won't run stateful tests because they were manually disabled.")
continue
# Reverse sort order: we want run newest test first.
# And not reverse subtests
def key_func(item):
if args.order == 'random':
return random()
reverse = 1 if args.order == 'asc' else -1
if -1 == item.find('_'):
return 99998
prefix, suffix = item.split('_', 1)
try:
return reverse * int(prefix), suffix
except ValueError:
return 99997
run_n, run_total = args.parallel.split('/')
run_n = float(run_n)
run_total = float(run_total)
all_tests = os.listdir(suite_dir)
all_tests = filter(lambda case: is_test_from_dir(suite_dir, case), all_tests)
all_tests = sorted(filter(lambda case: re.search(args.test, case) if args.test else True, all_tests), key=key_func)
tests_n = len(all_tests)
start = int(tests_n / run_total * (run_n - 1))
if start > 0:
start = start + 1
end = int(tests_n / run_total * (run_n))
all_tests = all_tests[start : end]
print("\nRunning {} {} tests.".format(tests_n, suite) + (" {} .. {} ".format(start, end) if run_total > 1 else "") + "\n")
if len(all_tests):
print("\nRunning {} {} tests.".format(len(all_tests), suite) + "\n")
for case in all_tests:
if SERVER_DIED:
@ -396,7 +281,7 @@ def main(args):
if os.path.exists(stderr_file):
os.remove(stderr_file)
except KeyboardInterrupt as e:
print(colored("Break tests execution", "red"))
print(colored("Break tests execution", args, "red"))
raise e
except:
import traceback
@ -414,25 +299,171 @@ def main(args):
failures_total = failures_total + failures
exit_code = 0
if failures_total > 0:
print(colored("\nHaving {failures_total} errors! {passed_total} tests passed. {skipped_total} tests skipped.".format(passed_total = passed_total, skipped_total = skipped_total, failures_total = failures_total), "red", attrs=["bold"]))
print(colored("\nHaving {failures_total} errors! {passed_total} tests passed. {skipped_total} tests skipped.".format(passed_total = passed_total, skipped_total = skipped_total, failures_total = failures_total), args, "red", attrs=["bold"]))
exit_code = 1
else:
print(colored("\n{passed_total} tests passed. {skipped_total} tests skipped.".format(passed_total = passed_total, skipped_total = skipped_total), "green", attrs=["bold"]))
print(colored("\n{passed_total} tests passed. {skipped_total} tests skipped.".format(passed_total = passed_total, skipped_total = skipped_total), args, "green", attrs=["bold"]))
server_logs_level = "warning"
def main(args):
global SERVER_DIED
global exit_code
global server_logs_level
def is_data_present():
clickhouse_proc = Popen(shlex.split(args.client), stdin=PIPE, stdout=PIPE, stderr=PIPE)
(stdout, stderr) = clickhouse_proc.communicate("EXISTS TABLE test.hits")
if clickhouse_proc.returncode != 0:
raise CalledProcessError(clickhouse_proc.returncode, args.client, stderr)
return stdout.startswith('1')
base_dir = os.path.abspath(args.queries)
tmp_dir = os.path.abspath(args.tmp)
# Keep same default values as in queries/shell_config.sh
os.environ.setdefault("CLICKHOUSE_BINARY", args.binary)
#os.environ.setdefault("CLICKHOUSE_CLIENT", args.client)
os.environ.setdefault("CLICKHOUSE_CONFIG", args.configserver)
if args.configclient:
os.environ.setdefault("CLICKHOUSE_CONFIG_CLIENT", args.configclient)
os.environ.setdefault("CLICKHOUSE_TMP", tmp_dir)
os.environ.setdefault("CLICKHOUSE_DATABASE", args.database)
# Force to print server warnings in stderr
# Shell scripts could change logging level
os.environ.setdefault("CLICKHOUSE_CLIENT_SERVER_LOGS_LEVEL", server_logs_level)
if args.zookeeper is None:
code, out = commands.getstatusoutput(args.extract_from_config +" --try --config " + args.configserver + ' --key zookeeper | grep . | wc -l')
try:
if int(out) > 0:
args.zookeeper = True
else:
args.zookeeper = False
except ValueError:
args.zookeeper = False
if args.shard is None:
code, out = commands.getstatusoutput(args.extract_from_config + " --try --config " + args.configserver + ' --key listen_host | grep -E "127.0.0.2|::"')
if out:
args.shard = True
else:
args.shard = False
clickhouse_proc_create = Popen(shlex.split(args.client), stdin=PIPE, stdout=PIPE, stderr=PIPE)
clickhouse_proc_create.communicate("CREATE DATABASE IF NOT EXISTS " + args.database)
if args.database != "test":
clickhouse_proc_create = Popen(shlex.split(args.client), stdin=PIPE, stdout=PIPE, stderr=PIPE)
clickhouse_proc_create.communicate("CREATE DATABASE IF NOT EXISTS test")
def is_test_from_dir(suite_dir, case):
case_file = os.path.join(suite_dir, case)
(name, ext) = os.path.splitext(case)
return os.path.isfile(case_file) and (ext == '.sql' or ext == '.sh' or ext == '.py')
def sute_key_func(item):
if args.order == 'random':
return random()
if -1 == item.find('_'):
return 99998
prefix, suffix = item.split('_', 1)
try:
return int(prefix), suffix
except ValueError:
return 99997
for suite in sorted(os.listdir(base_dir), key=sute_key_func):
if SERVER_DIED:
break
suite_dir = os.path.join(base_dir, suite)
suite_re_obj = re.search('^[0-9]+_(.*)$', suite)
if not suite_re_obj: #skip .gitignore and so on
continue
suite_tmp_dir = os.path.join(tmp_dir, suite)
if not os.path.exists(suite_tmp_dir):
os.makedirs(suite_tmp_dir)
suite = suite_re_obj.group(1)
if os.path.isdir(suite_dir):
if 'stateful' in suite and not args.no_stateful and not is_data_present():
print("Won't run stateful tests because test data wasn't loaded.")
continue
if 'stateless' in suite and args.no_stateless:
print("Won't run stateless tests because they were manually disabled.")
continue
if 'stateful' in suite and args.no_stateful:
print("Won't run stateful tests because they were manually disabled.")
continue
# Reverse sort order: we want run newest test first.
# And not reverse subtests
def key_func(item):
if args.order == 'random':
return random()
reverse = 1 if args.order == 'asc' else -1
if -1 == item.find('_'):
return 99998
prefix, suffix = item.split('_', 1)
try:
return reverse * int(prefix), suffix
except ValueError:
return 99997
all_tests = os.listdir(suite_dir)
all_tests = filter(lambda case: is_test_from_dir(suite_dir, case), all_tests)
all_tests = sorted(filter(lambda case: re.search(args.test, case) if args.test else True, all_tests), key=key_func)
run_n, run_total = args.parallel.split('/')
run_n = float(run_n)
run_total = float(run_total)
tests_n = len(all_tests)
if run_total > tests_n:
run_total = tests_n
if run_n > run_total:
continue
jobs = args.jobs
if jobs > run_total:
run_total = jobs
all_tests_array = []
for n in range(1, 1 + int(run_total)):
start = int(tests_n / run_total * (n - 1))
end = int(tests_n / run_total * n)
all_tests_array.append([all_tests[start : end], suite, suite_dir, suite_tmp_dir, run_total])
if jobs > 1:
with closing(Pool(processes=jobs)) as pool:
pool.map(run_tests_array, all_tests_array)
pool.terminate()
else:
run_tests_array(all_tests_array[int(run_n)-1])
if args.hung_check:
processlist = get_processlist(args.client_with_database)
if processlist:
server_pid = get_server_pid(os.getenv("CLICKHOUSE_PORT_TCP", '9000'))
print(colored("\nFound hung queries in processlist:", "red", attrs=["bold"]))
print(colored("\nFound hung queries in processlist:", args, "red", attrs=["bold"]))
print(processlist)
if server_pid:
print("\nStacktraces of all threads:")
print(get_stacktraces(server_pid))
exit_code = 1
else:
print(colored("\nNo queries hung.", "green", attrs=["bold"]))
print(colored("\nNo queries hung.", args, "green", attrs=["bold"]))
sys.exit(exit_code)
@ -467,7 +498,8 @@ if __name__ == '__main__':
parser.add_argument('--hung-check', action='store_true', default=False)
parser.add_argument('--force-color', action='store_true', default=False)
parser.add_argument('--database', default='test', help='Default database for tests')
parser.add_argument('--parallel', default='1/1', help='Parralel test run number/total')
parser.add_argument('--parallel', default='1/1', help='One parallel test run number/total')
parser.add_argument('-j', '--jobs', default=1, help='Run all tests in parallel', type=int)
parser.add_argument('--no-stateless', action='store_true', help='Disable all stateless tests')
parser.add_argument('--no-stateful', action='store_true', help='Disable all stateful tests')

2
dbms/tests/integration/helpers/docker_compose_kafka.yml
View File

@ -12,7 +12,7 @@ services:
- label:disable
kafka1:
image: confluentinc/cp-kafka:4.1.0
image: confluentinc/cp-kafka:5.2.0
hostname: kafka1
ports:
- "9092:9092"

127
dbms/tests/integration/test_storage_kafka/test.py
View File

@ -7,7 +7,8 @@ from helpers.test_tools import TSV
import json
import subprocess
from kafka import KafkaProducer
import kafka.errors
from kafka import KafkaAdminClient, KafkaProducer
from google.protobuf.internal.encoder import _VarintBytes
"""
@ -62,22 +63,11 @@ def wait_kafka_is_available(max_retries=50):
def kafka_produce(topic, messages):
p = subprocess.Popen(('docker',
'exec',
'-i',
kafka_id,
'/usr/bin/kafka-console-producer',
'--broker-list',
'INSIDE://localhost:9092',
'--topic',
topic,
'--sync',
'--message-send-max-retries',
'100'),
stdin=subprocess.PIPE)
p.communicate(messages)
p.stdin.close()
print("Produced {} messages for topic {}".format(len(messages.splitlines()), topic))
producer = KafkaProducer(bootstrap_servers="localhost:9092")
for message in messages:
producer.send(topic=topic, value=message)
producer.flush()
print ("Produced {} messages for topic {}".format(len(messages), topic))
def kafka_produce_protobuf_messages(topic, start_index, num_messages):
@ -141,9 +131,9 @@ def test_kafka_settings_old_syntax(kafka_cluster):
# Don't insert malformed messages since old settings syntax
# doesn't support skipping of broken messages.
messages = ''
messages = []
for i in range(50):
messages += json.dumps({'key': i, 'value': i}) + '\n'
messages.append(json.dumps({'key': i, 'value': i}))
kafka_produce('old', messages)
result = ''
@ -167,18 +157,18 @@ def test_kafka_settings_new_syntax(kafka_cluster):
kafka_skip_broken_messages = 1;
''')
messages = ''
messages = []
for i in range(25):
messages += json.dumps({'key': i, 'value': i}) + '\n'
messages.append(json.dumps({'key': i, 'value': i}))
kafka_produce('new', messages)
# Insert couple of malformed messages.
kafka_produce('new', '}{very_broken_message,\n')
kafka_produce('new', '}another{very_broken_message,\n')
kafka_produce('new', ['}{very_broken_message,'])
kafka_produce('new', ['}another{very_broken_message,'])
messages = ''
messages = []
for i in range(25, 50):
messages += json.dumps({'key': i, 'value': i}) + '\n'
messages.append(json.dumps({'key': i, 'value': i}))
kafka_produce('new', messages)
result = ''
@ -201,9 +191,9 @@ def test_kafka_csv_with_delimiter(kafka_cluster):
kafka_row_delimiter = '\\n';
''')
messages = ''
messages = []
for i in range(50):
messages += '{i}, {i}\n'.format(i=i)
messages.append('{i}, {i}'.format(i=i))
kafka_produce('csv', messages)
result = ''
@ -226,9 +216,9 @@ def test_kafka_tsv_with_delimiter(kafka_cluster):
kafka_row_delimiter = '\\n';
''')
messages = ''
messages = []
for i in range(50):
messages += '{i}\t{i}\n'.format(i=i)
messages.append('{i}\t{i}'.format(i=i))
kafka_produce('tsv', messages)
result = ''
@ -239,6 +229,35 @@ def test_kafka_tsv_with_delimiter(kafka_cluster):
kafka_check_result(result, True)
def test_kafka_json_without_delimiter(kafka_cluster):
instance.query('''
CREATE TABLE test.kafka (key UInt64, value UInt64)
ENGINE = Kafka
SETTINGS
kafka_broker_list = 'kafka1:19092',
kafka_topic_list = 'json',
kafka_group_name = 'json',
kafka_format = 'JSONEachRow';
''')
messages = ''
for i in range(25):
messages += json.dumps({'key': i, 'value': i}) + '\n'
kafka_produce('json', [messages])
messages = ''
for i in range(25, 50):
messages += json.dumps({'key': i, 'value': i}) + '\n'
kafka_produce('json', [messages])
result = ''
for i in range(50):
result += instance.query('SELECT * FROM test.kafka')
if kafka_check_result(result):
break
kafka_check_result(result, True)
def test_kafka_protobuf(kafka_cluster):
instance.query('''
CREATE TABLE test.kafka (key UInt64, value String)
@ -282,9 +301,9 @@ def test_kafka_materialized_view(kafka_cluster):
SELECT * FROM test.kafka;
''')
messages = ''
messages = []
for i in range(50):
messages += json.dumps({'key': i, 'value': i}) + '\n'
messages.append(json.dumps({'key': i, 'value': i}))
kafka_produce('json', messages)
for i in range(20):
@ -300,6 +319,52 @@ def test_kafka_materialized_view(kafka_cluster):
''')
def test_kafka_flush_on_big_message(kafka_cluster):
# Create batchs of messages of size ~100Kb
kafka_messages = 10000
batch_messages = 1000
messages = [json.dumps({'key': i, 'value': 'x' * 100}) * batch_messages for i in range(kafka_messages)]
kafka_produce('flush', messages)
instance.query('''
DROP TABLE IF EXISTS test.view;
DROP TABLE IF EXISTS test.consumer;
CREATE TABLE test.kafka (key UInt64, value String)
ENGINE = Kafka
SETTINGS
kafka_broker_list = 'kafka1:19092',
kafka_topic_list = 'flush',
kafka_group_name = 'flush',
kafka_format = 'JSONEachRow',
kafka_max_block_size = 10;
CREATE TABLE test.view (key UInt64, value String)
ENGINE = MergeTree
ORDER BY key;
CREATE MATERIALIZED VIEW test.consumer TO test.view AS
SELECT * FROM test.kafka;
''')
client = KafkaAdminClient(bootstrap_servers="localhost:9092")
received = False
while not received:
try:
offsets = client.list_consumer_group_offsets('flush')
for topic, offset in offsets.items():
if topic.topic == 'flush' and offset.offset == kafka_messages:
received = True
break
except kafka.errors.GroupCoordinatorNotAvailableError:
continue
for _ in range(20):
time.sleep(1)
result = instance.query('SELECT count() FROM test.view')
if int(result) == kafka_messages*batch_messages:
break
assert int(result) == kafka_messages*batch_messages, 'ClickHouse lost some messages: {}'.format(result)
if __name__ == '__main__':
cluster.start()
raw_input("Cluster created, press any key to destroy...")

19
dbms/tests/queries/0_stateless/00722_inner_join.reference
View File

@ -19,10 +19,21 @@
┌─database─┬─name─┐
│ system │ one │
└──────────┴──────┘
one
system one
system one
test one
┌─db.x───┬─name─┐
│ system │ one │
└────────┴──────┘
┌─db.name─┬─name─┐
│ system │ one │
└─────────┴──────┘
┌─db.name─┬─name─┐
│ system │ one │
└─────────┴──────┘
┌─database─┬─name─┐
│ system │ one │
└──────────┴──────┘
┌─database─┬─name─┐
│ system │ one │
└──────────┴──────┘
2
2
2

36
dbms/tests/queries/0_stateless/00722_inner_join.sql
View File

@ -1,6 +1,5 @@
CREATE DATABASE IF NOT EXISTS test;
DROP TABLE IF EXISTS test.one;
CREATE TABLE test.one(dummy UInt8) ENGINE = Memory;
DROP TABLE IF EXISTS one;
CREATE TABLE one(dummy UInt8) ENGINE = Memory;
SELECT database, t.name
FROM system.tables AS t
@ -48,25 +47,34 @@ SELECT database, t.name
WHERE db.name = 'system' AND t.name = 'one'
FORMAT PrettyCompactNoEscapes;
--SELECT db.name, t.name
-- FROM (SELECT name, database AS x FROM system.tables) AS t
-- JOIN (SELECT name AS x FROM system.databases) AS db USING x
-- WHERE x = 'system' AND t.name = 'one'
-- FORMAT PrettyCompactNoEscapes;
SELECT db.x, t.name
FROM (SELECT name, database AS x FROM system.tables) AS t
JOIN (SELECT name AS x FROM system.databases) AS db USING x
WHERE x = 'system' AND t.name = 'one'
FORMAT PrettyCompactNoEscapes;
SELECT t.name --, db.name
SELECT db.name, t.name
FROM (SELECT name, database FROM system.tables WHERE name = 'one') AS t
JOIN (SELECT name FROM system.databases WHERE name = 'system') AS db ON t.database = db.name;
JOIN (SELECT name FROM system.databases WHERE name = 'system') AS db ON t.database = db.name
FORMAT PrettyCompactNoEscapes;
SELECT db.name, t.name
FROM system.tables AS t
JOIN (SELECT * FROM system.databases WHERE name = 'system') AS db ON t.database = db.name
WHERE t.name = 'one';
WHERE t.name = 'one'
FORMAT PrettyCompactNoEscapes;
SELECT database, t.name
SELECT t.database, t.name
FROM system.tables AS t
JOIN (SELECT name, name AS database FROM system.databases) AS db ON t.database = db.name
WHERE t.name = 'one';
WHERE t.database = 'system' AND t.name = 'one'
FORMAT PrettyCompactNoEscapes;
SELECT t.database, t.name
FROM system.tables t
ANY LEFT JOIN (SELECT 'system' AS base, 'one' AS name) db USING name
WHERE t.database = db.base
FORMAT PrettyCompactNoEscapes;
SELECT count(t.database)
FROM (SELECT * FROM system.tables WHERE name = 'one') AS t
@ -82,4 +90,4 @@ SELECT count()
JOIN system.databases AS db ON db.name = t.database
WHERE t.name = 'one';
DROP TABLE test.one;
DROP TABLE one;

2
dbms/tests/queries/0_stateless/00860_unknown_identifier_bug.reference Normal file
View File

@ -0,0 +1,2 @@
1 Created 1
2 Created 2

39
dbms/tests/queries/0_stateless/00860_unknown_identifier_bug.sql Normal file
View File

@ -0,0 +1,39 @@
DROP TABLE IF EXISTS appointment_events;
CREATE TABLE appointment_events
(
_appointment_id UInt32,
_id String,
_status String,
_set_by_id String,
_company_id String,
_client_id String,
_type String,
_at String,
_vacancy_id String,
_set_at UInt32,
_job_requisition_id String
) ENGINE = Memory;
INSERT INTO appointment_events (_appointment_id, _set_at, _status) values (1, 1, 'Created'), (2, 2, 'Created');
SELECT A._appointment_id,
A._id,
A._status,
A._set_by_id,
A._company_id,
A._client_id,
A._type,
A._at,
A._vacancy_id,
A._set_at,
A._job_requisition_id
FROM appointment_events A ANY
LEFT JOIN
(SELECT _appointment_id,
MAX(_set_at) AS max_set_at
FROM appointment_events
WHERE _status in ('Created', 'Transferred')
GROUP BY _appointment_id ) B USING _appointment_id
WHERE A._set_at = B.max_set_at;
DROP TABLE appointment_events;

6
dbms/tests/queries/0_stateless/00926_zookeeper_adaptive_index_granularity_replicated_merge_tree.sql
View File

@ -35,7 +35,7 @@ SELECT distinct(marks) from system.parts WHERE table = 'zero_rows_per_granule2'
SELECT distinct(marks) from system.parts WHERE table = 'zero_rows_per_granule1' and database='test' and active=1;
SELECT sleep(0.5) Format Null;
SELECT sleep(0.7) Format Null;
OPTIMIZE TABLE test.zero_rows_per_granule2 FINAL;
@ -92,7 +92,7 @@ SELECT distinct(marks) from system.parts WHERE table = 'four_rows_per_granule2'
SELECT distinct(marks) from system.parts WHERE table = 'four_rows_per_granule1' and database='test' and active=1;
SELECT sleep(0.5) Format Null;
SELECT sleep(0.7) Format Null;
OPTIMIZE TABLE test.four_rows_per_granule2 FINAL;
@ -180,7 +180,7 @@ ATTACH TABLE test.adaptive_granularity_alter1;
INSERT INTO test.adaptive_granularity_alter1 (p, k, v1, v2) VALUES ('2018-05-15', 100, 1000, 'aaaa'), ('2018-05-16', 101, 3000, 'bbbb'), ('2018-05-17', 102, 5000, 'cccc'), ('2018-05-19', 103, 7000, 'dddd');
SELECT sleep(0.5) Format Null;
SELECT sleep(0.7) Format Null;
OPTIMIZE TABLE test.adaptive_granularity_alter1 FINAL;

19
dbms/tests/queries/0_stateless/00933_ttl_simple.sql
View File

@ -47,22 +47,3 @@ create table ttl_00933_1 (d DateTime, a Int ttl toDateTime(1)) engine = MergeTre
create table ttl_00933_1 (d DateTime, a Int ttl d - d) engine = MergeTree order by tuple() partition by toSecond(d); -- { serverError 450 }
drop table if exists ttl_00933_1;
/*
Alexey Milovidov, [17.04.19 20:09]
sleep(0.7)
sleep(1.1)
- почему? @Alesapin
Alexander Sapin, [17.04.19 23:16]
[In reply to Alexey Milovidov]
1.1 по логике теста, я попробовал с 0.5 и у меня флапнуло. С 1 не флапало, но работало долго. Попробовал 0.7 и тоже не флапает.
Alexey Milovidov, [17.04.19 23:18]
Слабо такой комментарий добавить прямо в тест? :)
Alexander Sapin, [17.04.19 23:20]
как-то неловко :)
*/

4
docker/test/stress/Dockerfile
View File

@ -18,7 +18,9 @@ RUN apt-get update -y \
openssl \
netcat-openbsd \
telnet \
llvm-8
llvm-8 \
brotli
COPY ./stress /stress
COPY log_queries.xml /etc/clickhouse-server/users.d/log_queries.xml

4
docker/test/stress/stress
View File

@ -11,7 +11,7 @@ import time
def run_perf_test(cmd, xmls_path, output_folder):
output_path = os.path.join(output_folder, "perf_stress_run.txt")
f = open(output_path, 'w')
p = Popen("{} --skip-tags=long --r {}".format(cmd, xmls_path), shell=True, stdout=f, stderr=f)
p = Popen("{} --skip-tags=long --recursive --input-files {}".format(cmd, xmls_path), shell=True, stdout=f, stderr=f)
return p
def run_func_test(cmd, output_prefix, num_processes):
@ -55,7 +55,7 @@ if __name__ == "__main__":
while True:
retcodes = []
for p in func_pipes:
if p.poll():
if p.poll() is not None:
retcodes.append(p.returncode)
if len(retcodes) == len(func_pipes):
break

2
docs/en/operations/table_engines/replication.md
View File

@ -55,7 +55,7 @@ For very large clusters, you can use different ZooKeeper clusters for different
Replication is asynchronous and multi-master. `INSERT` queries (as well as `ALTER`) can be sent to any available server. Data is inserted on the server where the query is run, and then it is copied to the other servers. Because it is asynchronous, recently inserted data appears on the other replicas with some latency. If part of the replicas are not available, the data is written when they become available. If a replica is available, the latency is the amount of time it takes to transfer the block of compressed data over the network.
By default, an INSERT query waits for confirmation of writing the data from only one replica. If the data was successfully written to only one replica and the server with this replica ceases to exist, the stored data will be lost. Tp enable getting confirmation of data writes from multiple replicas, use the `insert_quorum` option.
By default, an INSERT query waits for confirmation of writing the data from only one replica. If the data was successfully written to only one replica and the server with this replica ceases to exist, the stored data will be lost. To enable getting confirmation of data writes from multiple replicas, use the `insert_quorum` option.
Each block of data is written atomically. The INSERT query is divided into blocks up to `max_insert_block_size = 1048576` rows. In other words, if the `INSERT` query has less than 1048576 rows, it is made atomically.

2
docs/en/operations/table_engines/summingmergetree.md
View File

@ -86,7 +86,7 @@ SELECT key, sum(value) FROM summtt GROUP BY key
```
## Data Processing
## Data Processing {#data-processing}
When data are inserted into a table, they are saved as-is. Clickhouse merges the inserted parts of data periodically and this is when rows with the same primary key are summed and replaced with one for each resulting part of data.

66
docs/en/query_language/select.md
View File

@ -156,7 +156,7 @@ Here, a sample of 10% is taken from the second half of the data.
### ARRAY JOIN Clause {#select-array-join-clause}
Allows executing `JOIN` with an array or nested data structure. Allows you to perform `JOIN` both with the external array and with the inner array in the table. The intent is similar to the [arrayJoin](functions/array_functions.md#array_functions-join) function, but its functionality is broader.
Allows executing `JOIN` with an array or nested data structure. The intent is similar to the [arrayJoin](functions/array_join.md#functions_arrayjoin) function, but its functionality is broader.
``` sql
SELECT <expr_list>
@ -168,14 +168,14 @@ FROM <left_subquery>
You can specify only a single `ARRAY JOIN` clause in a query.
When running the `ARRAY JOIN`, there is an optimization of the query execution order. Although the `ARRAY JOIN` must be always specified before the `WHERE/PREWHERE` clause, it can be performed as before the `WHERE/PREWHERE` (if its result is needed in this clause), as after completing it (to reduce the volume of calculations). The processing order is controlled by the query optimizer.
The query execution order is optimized when running `ARRAY JOIN`. Although `ARRAY JOIN` must always be specified before the `WHERE/PREWHERE` clause, it can be performed either before `WHERE/PREWHERE` (if the result is needed in this clause), or after completing it (to reduce the volume of calculations). The processing order is controlled by the query optimizer.
Supported types of `ARRAY JOIN` are listed below:
- `ARRAY JOIN` - Executing `JOIN` with an array or nested data structure. Empty arrays are not included in the result.
- `LEFT ARRAY JOIN` - Unlike `ARRAY JOIN`, when using the `LEFT ARRAY JOIN` the result contains the rows with empty arrays. The value for an empty array is set to default value for an array element type (usually 0, empty string or NULL).
- `ARRAY JOIN` - In this case, empty arrays are not included in the result of `JOIN`.
- `LEFT ARRAY JOIN` - The result of `JOIN` contains rows with empty arrays. The value for an empty array is set to the default value for the array element type (usually 0, empty string or NULL).
Examples below demonstrate the usage of the `ARRAY JOIN` clause. Let's create a table with an [Array](../data_types/array.md) type column and insert values into it:
The examples below demonstrate the usage of the `ARRAY JOIN` and `LEFT ARRAY JOIN` clauses. Let's create a table with an [Array](../data_types/array.md) type column and insert values into it:
``` sql
CREATE TABLE arrays_test
@ -195,7 +195,7 @@ VALUES ('Hello', [1,2]), ('World', [3,4,5]), ('Goodbye', []);
└─────────────┴─────────┘
```
The first example shows using the `ARRAY JOIN` clause:
The example below uses the `ARRAY JOIN` clause:
``` sql
SELECT s, arr
@ -212,7 +212,7 @@ ARRAY JOIN arr;
└───────┴─────┘
```
The second example shows using the `LEFT ARRAY JOIN` clause:
The next example uses the `LEFT ARRAY JOIN` clause:
``` sql
SELECT s, arr
@ -230,7 +230,27 @@ LEFT ARRAY JOIN arr;
└─────────────┴─────┘
```
The next example demonstrates using the `ARRAY JOIN` with the external array:
#### Using Aliases
An alias can be specified for an array in the `ARRAY JOIN` clause. In this case, an array item can be accessed by this alias, but the array itself is accessed by the original name. Example:
``` sql
SELECT s, arr, a
FROM arrays_test
ARRAY JOIN arr AS a;
```
```
┌─s─────┬─arr─────┬─a─┐
│ Hello │ [1,2] │ 1 │
│ Hello │ [1,2] │ 2 │
│ World │ [3,4,5] │ 3 │
│ World │ [3,4,5] │ 4 │
│ World │ [3,4,5] │ 5 │
└───────┴─────────┴───┘
```
Using aliases, you can perform `ARRAY JOIN` with an external array. For example:
``` sql
SELECT s, arr_external
@ -252,27 +272,7 @@ ARRAY JOIN [1, 2, 3] AS arr_external;
└─────────────┴──────────────┘
```
#### Using Aliases
An alias can be specified for an array in the `ARRAY JOIN` clause. In this case, an array item can be accessed by this alias, but the array itself by the original name. Example:
``` sql
SELECT s, arr, a
FROM arrays_test
ARRAY JOIN arr AS a;
```
```
┌─s─────┬─arr─────┬─a─┐
│ Hello │ [1,2] │ 1 │
│ Hello │ [1,2] │ 2 │
│ World │ [3,4,5] │ 3 │
│ World │ [3,4,5] │ 4 │
│ World │ [3,4,5] │ 5 │
└───────┴─────────┴───┘
```
Multiple arrays of the same size can be comma-separated in the `ARRAY JOIN` clause. In this case, `JOIN` is performed with them simultaneously (the direct sum, not the cartesian product). Example:
Multiple arrays can be comma-separated in the `ARRAY JOIN` clause. In this case, `JOIN` is performed with them simultaneously (the direct sum, not the cartesian product). Note that all the arrays must have the same size. Example:
``` sql
SELECT s, arr, a, num, mapped
@ -290,6 +290,8 @@ ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num, arrayMap(x -> x + 1, arr) AS ma
└───────┴─────────┴───┴─────┴────────┘
```
The example below uses the [arrayEnumerate](functions/array_functions.md#array_functions-arrayenumerate) function:
``` sql
SELECT s, arr, a, num, arrayEnumerate(arr)
FROM arrays_test
@ -308,7 +310,7 @@ ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num;
#### ARRAY JOIN With Nested Data Structure
`ARRAY JOIN` also works with [nested data structure](../data_types/nested_data_structures/nested.md). Example:
`ARRAY `JOIN`` also works with [nested data structures](../data_types/nested_data_structures/nested.md). Example:
``` sql
CREATE TABLE nested_test
@ -401,7 +403,7 @@ ARRAY JOIN nest AS n;
└───────┴─────┴─────┴─────────┴────────────┘
```
The example of using the [arrayEnumerate](functions/array_functions.md#array_functions-arrayenumerate) function:
Example of using the [arrayEnumerate](functions/array_functions.md#array_functions-arrayenumerate) function:
``` sql
SELECT s, `n.x`, `n.y`, `nest.x`, `nest.y`, num
@ -444,7 +446,7 @@ The table names can be specified instead of `<left_subquery>` and `<right_subque
- `FULL JOIN` (or `FULL OUTER JOIN`)
- `CROSS JOIN` (or `,` )
See standard [SQL JOIN](https://en.wikipedia.org/wiki/Join_(SQL)) description.
See the standard [SQL JOIN](https://en.wikipedia.org/wiki/Join_(SQL)) description.
**ANY or ALL strictness**

2
docs/ru/operations/settings/settings.md
View File

@ -236,7 +236,7 @@ Ok.
## preferred_block_size_bytes
Служит для тех же целей что и `max_block_size`, но задает реккомедуемый размер блоков в байтах, выбирая адаптивное количество строк в блоке.
Служит для тех же целей что и `max_block_size`, но задает рекомедуемый размер блоков в байтах, выбирая адаптивное количество строк в блоке.
При этом размер блока не может быть более `max_block_size` строк.
По умолчанию: 1,000,000. Работает только при чтении из MergeTree-движков.

2
docs/ru/query_language/functions/array_functions.md
View File

@ -198,7 +198,7 @@ SELECT countEqual([1, 2, NULL, NULL], NULL)
└──────────────────────────────────────┘
```
## arrayEnumerate(arr)
## arrayEnumerate(arr) {#array_functions-arrayenumerate}
Возвращает массив \[1, 2, 3, ..., length(arr)\]

223
docs/ru/query_language/select.md
View File

@ -161,50 +161,53 @@ SAMPLE 1/10 OFFSET 1/2
### Секция ARRAY JOIN {#select-array-join-clause}
Позволяет выполнить JOIN с массивом или вложенной структурой данных. Смысл похож на функцию arrayJoin, но функциональность более широкая.
`ARRAY JOIN` - это, по сути, `INNER JOIN` с массивом. Пример:
Позволяет выполнить `JOIN` с массивом или вложенной структурой данных. Смысл похож на функцию [arrayJoin](functions/array_join.md#functions_arrayjoin), но функциональность более широкая.
``` sql
SELECT <expr_list>
FROM <left_subquery>
[LEFT] ARRAY JOIN <array>
[WHERE|PREWHERE <expr>]
...
```
:) CREATE TABLE arrays_test (s String, arr Array(UInt8)) ENGINE = Memory
В запросе может быть указано не более одной секции `ARRAY JOIN`.
При использовании `ARRAY JOIN`, порядок выполнения запроса оптимизируется. Несмотря на то что секция `ARRAY JOIN` всегда указывается перед выражением `WHERE / PREWHERE`, преобразование `JOIN` может быть выполнено как до выполнения выражения `WHERE / PREWHERE` (если результат необходим в этом выражении), так и после (чтобы уменьшить объем расчетов). Порядок обработки контролируется оптимизатором запросов.
Секция `ARRAY JOIN` поддерживает следующие формы записи:
- `ARRAY JOIN` — в этом случае результат `JOIN` не будет содержать пустые массивы;
- `LEFT ARRAY JOIN` — пустые массивы попадут в результат выполнения `JOIN`. В качестве значения для пустых массивов устанавливается значение по умолчанию. Обычно это 0, пустая строка или NULL, в зависимости от типа элементов массива.
Рассмотрим примеры использования `ARRAY JOIN` и `LEFT ARRAY JOIN`. Для начала создадим таблицу, содержащую столбец с типом [Array](../data_types/array.md), и добавим в него значение:
``` sql
CREATE TABLE arrays_test
(
s String,
arr Array(UInt8)
) ENGINE = Memory
) ENGINE = Memory;
Ok.
0 rows in set. Elapsed: 0.001 sec.
:) INSERT INTO arrays_test VALUES ('Hello', [1,2]), ('World', [3,4,5]), ('Goodbye', [])
INSERT INTO arrays_test VALUES
Ok.
3 rows in set. Elapsed: 0.001 sec.
:) SELECT * FROM arrays_test
SELECT *
FROM arrays_test
┌─s───────┬─arr─────┐
INSERT INTO arrays_test
VALUES ('Hello', [1,2]), ('World', [3,4,5]), ('Goodbye', []);
```
```
┌─s───────────┬─arr─────┐
│ Hello │ [1,2] │
│ World │ [3,4,5] │
│ Goodbye │ [] │
└─────────┴─────────┘
└─────────────┴─────────┘
```
3 rows in set. Elapsed: 0.001 sec.
:) SELECT s, arr FROM arrays_test ARRAY JOIN arr
В примере ниже используется `ARRAY JOIN`:
``` sql
SELECT s, arr
FROM arrays_test
ARRAY JOIN arr
ARRAY JOIN arr;
```
```
┌─s─────┬─arr─┐
│ Hello │ 1 │
│ Hello │ 2 │
@ -212,19 +215,37 @@ ARRAY JOIN arr
│ World │ 4 │
│ World │ 5 │
└───────┴─────┘
5 rows in set. Elapsed: 0.001 sec.
```
Для массива в секции ARRAY JOIN может быть указан алиас. В этом случае, элемент массива будет доступен под этим алиасом, а сам массив - под исходным именем. Пример:
Следующий пример использует `LEFT ARRAY JOIN`:
``` sql
SELECT s, arr
FROM arrays_test
LEFT ARRAY JOIN arr;
```
```
┌─s───────────┬─arr─┐
│ Hello │ 1 │
│ Hello │ 2 │
│ World │ 3 │
│ World │ 4 │
│ World │ 5 │
│ Goodbye │ 0 │
└─────────────┴─────┘
```
:) SELECT s, arr, a FROM arrays_test ARRAY JOIN arr AS a
#### Использование алиасов
Для массива в секции `ARRAY JOIN` может быть указан алиас. В этом случае, элемент массива будет доступен под этим алиасом, а сам массив — под исходным именем. Пример:
``` sql
SELECT s, arr, a
FROM arrays_test
ARRAY JOIN arr AS a
ARRAY JOIN arr AS a;
```
```
┌─s─────┬─arr─────┬─a─┐
│ Hello │ [1,2] │ 1 │
│ Hello │ [1,2] │ 2 │
@ -232,19 +253,39 @@ ARRAY JOIN arr AS a
│ World │ [3,4,5] │ 4 │
│ World │ [3,4,5] │ 5 │
└───────┴─────────┴───┘
5 rows in set. Elapsed: 0.001 sec.
```
В секции ARRAY JOIN может быть указано несколько массивов одинаковых размеров через запятую. В этом случае, JOIN делается с ними одновременно (прямая сумма, а не прямое произведение). Пример:
Используя алиасы, можно выполнять `JOIN` с внешними массивами:
``` sql
SELECT s, arr_external
FROM arrays_test
ARRAY JOIN [1, 2, 3] AS arr_external;
```
```
:) SELECT s, arr, a, num, mapped FROM arrays_test ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num, arrayMap(x -> x + 1, arr) AS mapped
┌─s───────────┬─arr_external─┐
│ Hello │ 1 │
│ Hello │ 2 │
│ Hello │ 3 │
│ World │ 1 │
│ World │ 2 │
│ World │ 3 │
│ Goodbye │ 1 │
│ Goodbye │ 2 │
│ Goodbye │ 3 │
└─────────────┴──────────────┘
```
В секции `ARRAY JOIN` можно указать через запятую сразу несколько массивов. В этом случае, `JOIN` делается с ними одновременно (прямая сумма, а не прямое произведение). Обратите внимание, массивы должны быть одинаковых размеров. Примеры:
``` sql
SELECT s, arr, a, num, mapped
FROM arrays_test
ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num, arrayMap(lambda(tuple(x), plus(x, 1)), arr) AS mapped
ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num, arrayMap(x -> x + 1, arr) AS mapped;
```
```
┌─s─────┬─arr─────┬─a─┬─num─┬─mapped─┐
│ Hello │ [1,2] │ 1 │ 1 │ 2 │
│ Hello │ [1,2] │ 2 │ 2 │ 3 │
@ -252,15 +293,17 @@ ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num, arrayMap(lambda(tuple(x), plus(
│ World │ [3,4,5] │ 4 │ 2 │ 5 │
│ World │ [3,4,5] │ 5 │ 3 │ 6 │
└───────┴─────────┴───┴─────┴────────┘
```
5 rows in set. Elapsed: 0.002 sec.
:) SELECT s, arr, a, num, arrayEnumerate(arr) FROM arrays_test ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num
В примере ниже используется функция [arrayEnumerate](functions/array_functions.md#array_functions-arrayenumerate):
``` sql
SELECT s, arr, a, num, arrayEnumerate(arr)
FROM arrays_test
ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num
ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num;
```
```
┌─s─────┬─arr─────┬─a─┬─num─┬─arrayEnumerate(arr)─┐
│ Hello │ [1,2] │ 1 │ 1 │ [1,2] │
│ Hello │ [1,2] │ 2 │ 2 │ [1,2] │
@ -268,54 +311,40 @@ ARRAY JOIN arr AS a, arrayEnumerate(arr) AS num
│ World │ [3,4,5] │ 4 │ 2 │ [1,2,3] │
│ World │ [3,4,5] │ 5 │ 3 │ [1,2,3] │
└───────┴─────────┴───┴─────┴─────────────────────┘
5 rows in set. Elapsed: 0.002 sec.
```
ARRAY JOIN также работает с вложенными структурами данных. Пример:
#### ARRAY JOIN с вложенными структурами данных
```
:) CREATE TABLE nested_test (s String, nest Nested(x UInt8, y UInt32)) ENGINE = Memory
`ARRAY JOIN` также работает с [вложенными структурами данных](../data_types/nested_data_structures/nested.md). Пример:
``` sql
CREATE TABLE nested_test
(
s String,
nest Nested(
x UInt8,
y UInt32)
) ENGINE = Memory
) ENGINE = Memory;
Ok.
0 rows in set. Elapsed: 0.006 sec.
:) INSERT INTO nested_test VALUES ('Hello', [1,2], [10,20]), ('World', [3,4,5], [30,40,50]), ('Goodbye', [], [])
INSERT INTO nested_test VALUES
Ok.
3 rows in set. Elapsed: 0.001 sec.
:) SELECT * FROM nested_test
SELECT *
FROM nested_test
INSERT INTO nested_test
VALUES ('Hello', [1,2], [10,20]), ('World', [3,4,5], [30,40,50]), ('Goodbye', [], []);
```
```
┌─s───────┬─nest.x──┬─nest.y─────┐
│ Hello │ [1,2] │ [10,20] │
│ World │ [3,4,5] │ [30,40,50] │
│ Goodbye │ [] │ [] │
└─────────┴─────────┴────────────┘
```
3 rows in set. Elapsed: 0.001 sec.
:) SELECT s, nest.x, nest.y FROM nested_test ARRAY JOIN nest
``` sql
SELECT s, `nest.x`, `nest.y`
FROM nested_test
ARRAY JOIN nest
ARRAY JOIN nest;
```
```
┌─s─────┬─nest.x─┬─nest.y─┐
│ Hello │ 1 │ 10 │
│ Hello │ 2 │ 20 │
@ -323,19 +352,17 @@ ARRAY JOIN nest
│ World │ 4 │ 40 │
│ World │ 5 │ 50 │
└───────┴────────┴────────┘
5 rows in set. Elapsed: 0.001 sec.
```
При указании имени вложенной структуры данных в ARRAY JOIN, смысл такой же, как ARRAY JOIN со всеми элементами-массивами, из которых она состоит. Пример:
```
:) SELECT s, nest.x, nest.y FROM nested_test ARRAY JOIN nest.x, nest.y
При указании имени вложенной структуры данных в `ARRAY JOIN`, смысл такой же, как `ARRAY JOIN` со всеми элементами-массивами, из которых она состоит. Пример:
``` sql
SELECT s, `nest.x`, `nest.y`
FROM nested_test
ARRAY JOIN `nest.x`, `nest.y`
ARRAY JOIN `nest.x`, `nest.y`;
```
```
┌─s─────┬─nest.x─┬─nest.y─┐
│ Hello │ 1 │ 10 │
│ Hello │ 2 │ 20 │
@ -343,19 +370,17 @@ ARRAY JOIN `nest.x`, `nest.y`
│ World │ 4 │ 40 │
│ World │ 5 │ 50 │
└───────┴────────┴────────┘
5 rows in set. Elapsed: 0.001 sec.
```
Такой вариант тоже имеет смысл:
```
:) SELECT s, nest.x, nest.y FROM nested_test ARRAY JOIN nest.x
``` sql
SELECT s, `nest.x`, `nest.y`
FROM nested_test
ARRAY JOIN `nest.x`
ARRAY JOIN `nest.x`;
```
```
┌─s─────┬─nest.x─┬─nest.y─────┐
│ Hello │ 1 │ [10,20] │
│ Hello │ 2 │ [10,20] │
@ -363,19 +388,17 @@ ARRAY JOIN `nest.x`
│ World │ 4 │ [30,40,50] │
│ World │ 5 │ [30,40,50] │
└───────┴────────┴────────────┘
5 rows in set. Elapsed: 0.001 sec.
```
Алиас для вложенной структуры данных можно использовать, чтобы выбрать как результат JOIN-а, так и исходный массив. Пример:
```
:) SELECT s, n.x, n.y, nest.x, nest.y FROM nested_test ARRAY JOIN nest AS n
Алиас для вложенной структуры данных можно использовать, чтобы выбрать как результат `JOIN`-а, так и исходный массив. Пример:
``` sql
SELECT s, `n.x`, `n.y`, `nest.x`, `nest.y`
FROM nested_test
ARRAY JOIN nest AS n
ARRAY JOIN nest AS n;
```
```
┌─s─────┬─n.x─┬─n.y─┬─nest.x──┬─nest.y─────┐
│ Hello │ 1 │ 10 │ [1,2] │ [10,20] │
│ Hello │ 2 │ 20 │ [1,2] │ [10,20] │
@ -383,19 +406,17 @@ ARRAY JOIN nest AS n
│ World │ 4 │ 40 │ [3,4,5] │ [30,40,50] │
│ World │ 5 │ 50 │ [3,4,5] │ [30,40,50] │
└───────┴─────┴─────┴─────────┴────────────┘
5 rows in set. Elapsed: 0.001 sec.
```
Пример использования функции arrayEnumerate:
```
:) SELECT s, n.x, n.y, nest.x, nest.y, num FROM nested_test ARRAY JOIN nest AS n, arrayEnumerate(nest.x) AS num
Пример использования функции [arrayEnumerate](functions/array_functions.md#array_functions-arrayenumerate):
``` sql
SELECT s, `n.x`, `n.y`, `nest.x`, `nest.y`, num
FROM nested_test
ARRAY JOIN nest AS n, arrayEnumerate(`nest.x`) AS num
ARRAY JOIN nest AS n, arrayEnumerate(`nest.x`) AS num;
```
```
┌─s─────┬─n.x─┬─n.y─┬─nest.x──┬─nest.y─────┬─num─┐
│ Hello │ 1 │ 10 │ [1,2] │ [10,20] │ 1 │
│ Hello │ 2 │ 20 │ [1,2] │ [10,20] │ 2 │
@ -403,16 +424,8 @@ ARRAY JOIN nest AS n, arrayEnumerate(`nest.x`) AS num
│ World │ 4 │ 40 │ [3,4,5] │ [30,40,50] │ 2 │
│ World │ 5 │ 50 │ [3,4,5] │ [30,40,50] │ 3 │
└───────┴─────┴─────┴─────────┴────────────┴─────┘
5 rows in set. Elapsed: 0.002 sec.
```
В запросе может быть указано не более одной секции ARRAY JOIN.
Соответствующее преобразование может выполняться как до секции WHERE/PREWHERE (если его результат нужен в этой секции), так и после выполнения WHERE/PREWHERE (чтобы уменьшить объём вычислений).
### Секция JOIN {#select-join}
Соединяет данные в привычном для [SQL JOIN](https://en.wikipedia.org/wiki/Join_(SQL)) смысле.

18
docs/zh/operations/table_engines/dictionary.md
View File

@ -1,9 +1,9 @@
# Dictionary
The `Dictionary` engine displays the dictionary data as a ClickHouse table.
`Dictionary` 引擎将字典数据展示为一个ClickHouse的表。
As an example, consider a dictionary of `products` with the following configuration:
例如，考虑使用一个具有以下配置的 `products` 字典：
```xml
<dictionaries>
@ -36,7 +36,7 @@ As an example, consider a dictionary of `products` with the following configurat
</dictionaries>
```
Query the dictionary data:
查询字典中的数据：
``` sql
select name, type, key, attribute.names, attribute.types, bytes_allocated, element_count,source from system.dictionaries where name = 'products';
@ -60,17 +60,17 @@ WHERE name = 'products'
└──────────┴──────┴────────┴─────────────────┴─────────────────┴─────────────────┴───────────────┴─────────────────┘
```
You can use the [dictGet*](../../query_language/functions/ext_dict_functions.md#ext_dict_functions) function to get the dictionary data in this format.
你可以使用 [dictGet*](../../query_language/functions/ext_dict_functions.md#ext_dict_functions) 函数来获取这种格式的字典数据。
This view isn't helpful when you need to get raw data, or when performing a `JOIN` operation. For these cases, you can use the `Dictionary` engine, which displays the dictionary data in a table.
当你需要获取原始数据，或者是想要使用 `JOIN` 操作的时候，这种视图并没有什么帮助。对于这些情况，你可以使用 `Dictionary` 引擎，它可以将字典数据展示在表中。
Syntax:
语法：
```
CREATE TABLE %table_name% (%fields%) engine = Dictionary(%dictionary_name%)`
```
Usage example:
示例：
``` sql
create table products (product_id UInt64, title String) Engine = Dictionary(products);
@ -89,7 +89,7 @@ Ok.
0 rows in set. Elapsed: 0.004 sec.
```
Take a look at what's in the table.
看一看表中的内容。
``` sql
select * from products limit 1;
@ -108,4 +108,4 @@ LIMIT 1
```
[Original article](https://clickhouse.yandex/docs/en/operations/table_engines/dictionary/) <!--hide-->
[来源文章](https://clickhouse.yandex/docs/en/operations/table_engines/dictionary/) <!--hide-->

34
docs/zh/operations/table_engines/mergetree.md
View File

@ -1,8 +1,8 @@
# MergeTree {#table_engines-mergetree}
Clickhouse 中最强大的表引擎当属 `MergeTree` （合并树）引擎及该家族（`*MergeTree`）中的其他引擎。
Clickhouse 中最强大的表引擎当属 `MergeTree` （合并树）引擎及该系列（`*MergeTree`）中的其他引擎。
`MergeTree` 引擎家族的基本理念如下。当你有巨量数据要插入到表中，你要高效地一批批写入数据分片，并希望这些数据分片在后台按照一定规则合并。相比在插入时不断修改（重写）数据进存储，这种策略会高效很多。
`MergeTree` 引擎系列的基本理念如下。当你有巨量数据要插入到表中，你要高效地一批批写入数据片段，并希望这些数据片段在后台按照一定规则合并。相比在插入时不断修改（重写）数据进存储，这种策略会高效很多。
主要特点:
@ -16,14 +16,14 @@ Clickhouse 中最强大的表引擎当属 `MergeTree` （合并树）引擎及
- 支持数据副本。
`ReplicatedMergeTree` 家族的表便是用于此。更多信息，请参阅 [数据副本](replication.md) 一节。
`ReplicatedMergeTree` 系列的表便是用于此。更多信息，请参阅 [数据副本](replication.md) 一节。
- 支持数据采样。
需要的话，你可以给表设置一个采样方法。
!!! 注意
[Merge](merge.md) 引擎并不属于 `*MergeTree` 家族。
[Merge](merge.md) 引擎并不属于 `*MergeTree` 系列。
## 建表 {#table_engine-mergetree-creating-a-table}
@ -70,8 +70,8 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
- `SETTINGS` — 影响 `MergeTree` 性能的额外参数：
- `index_granularity` — 索引粒度。即索引中相邻『标记』间的数据行数。默认值，8192 。该列表中所有可用的参数可以从这里查看 [MergeTreeSettings.h](https://github.com/yandex/ClickHouse/blob/master/dbms/src/Storages/MergeTree/MergeTreeSettings.h) 。
- `use_minimalistic_part_header_in_zookeeper` — 数据分片头在 ZooKeeper 中的存储方式。如果设置了 `use_minimalistic_part_header_in_zookeeper=1` ，ZooKeeper 会存储更少的数据。更多信息参考『服务配置参数』这章中的 [设置描述](../server_settings/settings.md#server-settings-use_minimalistic_part_header_in_zookeeper) 。
- `min_merge_bytes_to_use_direct_io` — 使用直接 I/O 来操作磁盘的合并操作时要求的最小数据量。合并数据分片时，ClickHouse 会计算要被合并的所有数据的总存储空间。如果大小超过了 `min_merge_bytes_to_use_direct_io` 设置的字节数，则 ClickHouse 将使用直接 I/O 接口（`O_DIRECT` 选项）对磁盘读写。如果设置 `min_merge_bytes_to_use_direct_io = 0` ，则会禁用直接 I/O。默认值：`10 * 1024 * 1024 * 1024` 字节。
- `use_minimalistic_part_header_in_zookeeper` — 数据片段头在 ZooKeeper 中的存储方式。如果设置了 `use_minimalistic_part_header_in_zookeeper=1` ，ZooKeeper 会存储更少的数据。更多信息参考『服务配置参数』这章中的 [设置描述](../server_settings/settings.md#server-settings-use_minimalistic_part_header_in_zookeeper) 。
- `min_merge_bytes_to_use_direct_io` — 使用直接 I/O 来操作磁盘的合并操作时要求的最小数据量。合并数据片段时，ClickHouse 会计算要被合并的所有数据的总存储空间。如果大小超过了 `min_merge_bytes_to_use_direct_io` 设置的字节数，则 ClickHouse 将使用直接 I/O 接口（`O_DIRECT` 选项）对磁盘读写。如果设置 `min_merge_bytes_to_use_direct_io = 0` ，则会禁用直接 I/O。默认值：`10 * 1024 * 1024 * 1024` 字节。
**示例配置**
@ -87,7 +87,7 @@ ENGINE MergeTree() PARTITION BY toYYYYMM(EventDate) ORDER BY (CounterID, EventDa
<details markdown="1"><summary>已弃用的建表方法</summary>
!!! 注意
!!! attention "注意"
不要在新版项目中使用该方法，可能的话，请将旧项目切换到上述方法。
```
@ -117,17 +117,17 @@ MergeTree(EventDate, intHash32(UserID), (CounterID, EventDate, intHash32(UserID)
## 数据存储
表由按主键排序的数据 *分片* 组成。
表由按主键排序的数据 *片段* 组成。
当数据被插入到表中时，会分成数据分片并按主键的字典序排序。例如，主键是 `(CounterID, Date)` 时，分片中数据按 `CounterID` 排序，具有相同 `CounterID` 的部分按 `Date` 排序。
当数据被插入到表中时，会分成数据片段并按主键的字典序排序。例如，主键是 `(CounterID, Date)` 时，片段中数据按 `CounterID` 排序，具有相同 `CounterID` 的部分按 `Date` 排序。
不同分区的数据会被分成不同的分片，ClickHouse 在后台合并数据分片以便更高效存储。不会合并来自不同分区的数据分片。这个合并机制并不保证相同主键的所有行都会合并到同一个数据分片中。
不同分区的数据会被分成不同的片段，ClickHouse 在后台合并数据片段以便更高效存储。不会合并来自不同分区的数据片段。这个合并机制并不保证相同主键的所有行都会合并到同一个数据片段中。
ClickHouse 会为每个数据分片创建一个索引文件，索引文件包含每个索引行（『标记』）的主键值。索引行号定义为 `n * index_granularity` 。最大的 `n` 等于总行数除以 `index_granularity` 的值的整数部分。对于每列，跟主键相同的索引行处也会写入『标记』。这些『标记』让你可以直接找到数据所在的列。
ClickHouse 会为每个数据片段创建一个索引文件，索引文件包含每个索引行（『标记』）的主键值。索引行号定义为 `n * index_granularity` 。最大的 `n` 等于总行数除以 `index_granularity` 的值的整数部分。对于每列，跟主键相同的索引行处也会写入『标记』。这些『标记』让你可以直接找到数据所在的列。
你可以只用一单一大表并不断地一块块往里面加入数据 – `MergeTree` 引擎的就是为了这样的场景。
## 主键和索引在查询中的表现 {#primary-keys-and-indexes-in-queriesko
## 主键和索引在查询中的表现 {#primary-keys-and-indexes-in-queries}
我们以 `(CounterID, Date)` 以主键。排序好的索引的图示会是下面这样：
@ -166,7 +166,7 @@ ClickHouse 不要求主键惟一。所以，你可以插入多条具有相同主
- 改善数据压缩。
ClickHouse 以主键排序分片数据，所以，数据的一致性越高，压缩越好。
ClickHouse 以主键排序片段数据，所以，数据的一致性越高，压缩越好。
- [CollapsingMergeTree](collapsingmergetree.md#table_engine-collapsingmergetree) 和 [SummingMergeTree](summingmergetree.md) 引擎里，数据合并时，会有额外的处理逻辑。
@ -177,7 +177,7 @@ ClickHouse 不要求主键惟一。所以，你可以插入多条具有相同主
### 选择跟排序键不一样主键
指定一个跟排序键（用于排序数据分片中行的表达式）
指定一个跟排序键（用于排序数据片段中行的表达式）
不一样的主键（用于计算写到索引文件的每个标记值的表达式）是可以的。
这种情况下，主键表达式元组必须是排序键表达式元组的一个前缀。
@ -192,7 +192,7 @@ ClickHouse 不要求主键惟一。所以，你可以插入多条具有相同主
这种情况下，主键中仅预留少量列保证高效范围扫描，
剩下的维度列放到排序键元组里。这样是合理的。
[排序键的修改](../../query_language/alter.md) 是轻量级的操作，因为一个新列同时被加入到表里和排序键后时，已存在的数据分片并不需要修改。由于旧的排序键是新排序键的前缀，并且刚刚添加的列中没有数据，因此在表修改时的数据对于新旧的排序键来说都是有序的。
[排序键的修改](../../query_language/alter.md) 是轻量级的操作，因为一个新列同时被加入到表里和排序键后时，已存在的数据片段并不需要修改。由于旧的排序键是新排序键的前缀，并且刚刚添加的列中没有数据，因此在表修改时的数据对于新旧的排序键来说都是有序的。
### 索引和分区在查询中的应用
@ -238,7 +238,7 @@ SELECT count() FROM table WHERE CounterID = 34 OR URL LIKE '%upyachka%'
INDEX index_name expr TYPE type(...) GRANULARITY granularity_value
```
`*MergeTree` 家族的表都能指定跳数索引。
`*MergeTree` 系列的表都能指定跳数索引。
这些索引是由数据块按粒度分割后的每部分在指定表达式上汇总信息 `granularity_value` 组成（粒度大小用表引擎里 `index_granularity` 的指定）。
这些汇总信息有助于用 `where` 语句跳过大片不满足的数据，从而减少 `SELECT` 查询从磁盘读取的数据量，
@ -292,7 +292,7 @@ INDEX sample_index3 (lower(str), str) TYPE ngrambf_v1(3, 256, 2, 0) GRANULARITY
## 并发数据访问
应对表的并发访问，我们使用多版本机制。换言之，当同时读和更新表时，数据从当前查询到的一组分片中读取。没有冗长的的锁。插入不会阻碍读取。
应对表的并发访问，我们使用多版本机制。换言之，当同时读和更新表时，数据从当前查询到的一组片段中读取。没有冗长的的锁。插入不会阻碍读取。
对表的读操作是自动并行的。

34
docs/zh/operations/table_engines/replacingmergetree.md
View File

@ -1,12 +1,12 @@
# ReplacingMergeTree
The engine differs from [MergeTree](mergetree.md) in that it removes duplicate entries with the same primary key value.
该引擎和[MergeTree](mergetree.md)的不同之处在于它会删除具有相同主键的重复项。
Data deduplication occurs only during a merge. Merging occurs in the background at an unknown time, so you can't plan for it. Some of the data may remain unprocessed. Although you can run an unscheduled merge using the `OPTIMIZE` query, don't count on using it, because the `OPTIMIZE` query will read and write a large amount of data.
数据的去重只会在合并的过程中出现。合并会在未知的时间在后台进行，因此你无法预先作出计划。有一些数据可能仍未被处理。尽管你可以调用 `OPTIMIZE` 语句发起计划外的合并，但请不要指望使用它，因为 `OPTIMIZE` 语句会引发对大量数据的读和写。
Thus, `ReplacingMergeTree` is suitable for clearing out duplicate data in the background in order to save space, but it doesn't guarantee the absence of duplicates.
因此，`ReplacingMergeTree` 适用于在后台清除重复的数据以节省空间，但是它不保证没有重复的数据出现。
## Creating a Table
## 建表
```sql
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
@ -21,24 +21,24 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
[SETTINGS name=value, ...]
```
For a description of request parameters, see [request description](../../query_language/create.md).
请求参数的描述，参考[请求参数](../../query_language/create.md)。
**ReplacingMergeTree Parameters**
- `ver` — column with version. Type `UInt*`, `Date` or `DateTime`. Optional parameter.
- `ver` — 版本列。类型为 `UInt*`, `Date` 或 `DateTime`。可选参数。
When merging, `ReplacingMergeTree` from all the rows with the same primary key leaves only one:
- Last in the selection, if `ver` not set.
- With the maximum version, if `ver` specified.
合并的时候，`ReplacingMergeTree` 从所有具有相同主键的行中选择一行留下：
- 如果 `ver` 列未指定，选择最后一条。
- 如果 `ver` 列已指定，选择 `ver` 值最大的版本。
**Query clauses**
**子句**
When creating a `ReplacingMergeTree` table the same [clauses](mergetree.md) are required, as when creating a `MergeTree` table.
创建 `ReplacingMergeTree` 表时，需要与创建 `MergeTree` 表时相同的[子句](mergetree.md)。
<details markdown="1"><summary>Deprecated Method for Creating a Table</summary>
<details markdown="1"><summary>已弃用的建表方法</summary>
!!! attention
Do not use this method in new projects and, if possible, switch the old projects to the method described above.
!!! attention "注意"
不要在新项目中使用该方法，可能的话，请将旧项目切换到上述方法。
```sql
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
@ -49,10 +49,10 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
) ENGINE [=] ReplacingMergeTree(date-column [, sampling_expression], (primary, key), index_granularity, [ver])
```
All of the parameters excepting `ver` have the same meaning as in `MergeTree`.
除了 `ver` 的所有参数都与 `MergeTree` 中的含义相同。
- `ver` - column with the version. Optional parameter. For a description, see the text above.
- `ver` - 版本列。可选参数，有关说明，请参阅上文。
</details>
[Original article](https://clickhouse.yandex/docs/en/operations/table_engines/replacingmergetree/) <!--hide-->
[来源文章](https://clickhouse.yandex/docs/en/operations/table_engines/replacingmergetree/) <!--hide-->

155
docs/zh/operations/table_engines/replication.md
View File

@ -1,27 +1,28 @@
# Data Replication {#table_engines-replication}
# 数据副本 {#table_engines-replication}
Replication is only supported for tables in the MergeTree family:
只有 MergeTree 系列里的表可支持副本：
- ReplicatedMergeTree
- ReplicatedSummingMergeTree
- ReplicatedReplacingMergeTree
- ReplicatedAggregatingMergeTree
- ReplicatedCollapsingMergeTree
- ReplicatedVersionedCollapsingMergeTree
- ReplicatedGraphiteMergeTree
Replication works at the level of an individual table, not the entire server. A server can store both replicated and non-replicated tables at the same time.
副本是表级别的，不是整个服务器级的。所以，服务器里可以同时有复制表和非复制表。
Replication does not depend on sharding. Each shard has its own independent replication.
副本不依赖分片。每个分片有它自己的独立副本。
Compressed data for `INSERT` and `ALTER` queries is replicated (for more information, see the documentation for [ALTER](../../query_language/alter.md#query_language_queries_alter)).
对于 `INSERT` 和 `ALTER` 语句操作数据的会在压缩的情况下被复制（更多信息，看 [ALTER](../../query_language/alter.md#query_language_queries_alter) ）。
`CREATE`, `DROP`, `ATTACH`, `DETACH` and `RENAME` queries are executed on a single server and are not replicated:
而 `CREATE`，`DROP`，`ATTACH`，`DETACH` 和 `RENAME` 语句只会在单个服务器上执行，不会被复制。
- `The CREATE TABLE` query creates a new replicatable table on the server where the query is run. If this table already exists on other servers, it adds a new replica.
- `The DROP TABLE` query deletes the replica located on the server where the query is run.
- `The RENAME` query renames the table on one of the replicas. In other words, replicated tables can have different names on different replicas.
- `The CREATE TABLE` 在运行此语句的服务器上创建一个新的可复制表。如果此表已存在其他服务器上，则给该表添加新副本。
- `The DROP TABLE` 删除运行此查询的服务器上的副本。
- `The RENAME` 重命名一个副本。换句话说，可复制表不同的副本可以有不同的名称。
To use replication, set the addresses of the ZooKeeper cluster in the config file. Example:
要使用副本，需在配置文件中设置 ZooKeeper 集群的地址。例如：
```xml
<zookeeper>
@ -40,43 +41,44 @@ To use replication, set the addresses of the ZooKeeper cluster in the config fil
</zookeeper>
```
Use ZooKeeper version 3.4.5 or later.
需要 ZooKeeper 3.4.5 或更高版本。
You can specify any existing ZooKeeper cluster and the system will use a directory on it for its own data (the directory is specified when creating a replicatable table).
你可以配置任何现有的 ZooKeeper 集群，系统会使用里面的目录来存取元数据（该目录在创建可复制表时指定）。
If ZooKeeper isn't set in the config file, you can't create replicated tables, and any existing replicated tables will be read-only.
如果配置文件中没有设置 ZooKeeper ，则无法创建复制表，并且任何现有的复制表都将变为只读。
ZooKeeper is not used in `SELECT` queries because replication does not affect the performance of `SELECT` and queries run just as fast as they do for non-replicated tables. When querying distributed replicated tables, ClickHouse behavior is controlled by the settings [max_replica_delay_for_distributed_queries](../settings/settings.md#settings-max_replica_delay_for_distributed_queries) and [fallback_to_stale_replicas_for_distributed_queries](../settings/settings.md).
`SELECT` 查询并不需要借助 ZooKeeper ，复本并不影响 `SELECT` 的性能，查询复制表与非复制表速度是一样的。查询分布式表时，ClickHouse的处理方式可通过设置 [max_replica_delay_for_distributed_queries](../settings/settings.md#settings-max_replica_delay_for_distributed_queries) 和 [fallback_to_stale_replicas_for_distributed_queries](../settings/settings.md) 修改。
For each `INSERT` query, approximately ten entries are added to ZooKeeper through several transactions. (To be more precise, this is for each inserted block of data; an INSERT query contains one block or one block per `max_insert_block_size = 1048576` rows.) This leads to slightly longer latencies for `INSERT` compared to non-replicated tables. But if you follow the recommendations to insert data in batches of no more than one `INSERT` per second, it doesn't create any problems. The entire ClickHouse cluster used for coordinating one ZooKeeper cluster has a total of several hundred `INSERTs` per second. The throughput on data inserts (the number of rows per second) is just as high as for non-replicated data.
对于每个 `INSERT` 语句，会通过几个事务将十来个记录添加到 ZooKeeper。（确切地说，这是针对每个插入的数据块; 每个 INSERT 语句的每 `max_insert_block_size = 1048576` 行和最后剩余的都各算作一个块。）相比非复制表，写 zk 会导致 `INSERT` 的延迟略长一些。但只要你按照建议每秒不超过一个 `INSERT` 地批量插入数据，不会有任何问题。一个 ZooKeeper 集群能给整个 ClickHouse 集群支撑协调每秒几百个 `INSERT`。数据插入的吞吐量（每秒的行数）可以跟不用复制的数据一样高。
For very large clusters, you can use different ZooKeeper clusters for different shards. However, this hasn't proven necessary on the Yandex.Metrica cluster (approximately 300 servers).
对于非常大的集群，你可以把不同的 ZooKeeper 集群用于不同的分片。然而，即使 Yandex.Metrica 集群（大约300台服务器）也证明还不需要这么做。
Replication is asynchronous and multi-master. `INSERT` queries (as well as `ALTER`) can be sent to any available server. Data is inserted on the server where the query is run, and then it is copied to the other servers. Because it is asynchronous, recently inserted data appears on the other replicas with some latency. If part of the replicas are not available, the data is written when they become available. If a replica is available, the latency is the amount of time it takes to transfer the block of compressed data over the network.
复制是多主异步。 `INSERT` 语句（以及 `ALTER` ）可以发给任意可用的服务器。数据会先插入到执行该语句的服务器上，然后被复制到其他服务器。由于它是异步的，在其他副本上最近插入的数据会有一些延迟。如果部分副本不可用，则数据在其可用时再写入。副本可用的情况下，则延迟时长是通过网络传输压缩数据块所需的时间。
By default, an INSERT query waits for confirmation of writing the data from only one replica. If the data was successfully written to only one replica and the server with this replica ceases to exist, the stored data will be lost. Tp enable getting confirmation of data writes from multiple replicas, use the `insert_quorum` option.
默认情况下，INSERT 语句仅等待一个副本写入成功后返回。如果数据只成功写入一个副本后该副本所在的服务器不再存在，则存储的数据会丢失。要启用数据写入多个副本才确认返回，使用 `insert_quorum` 选项。
Each block of data is written atomically. The INSERT query is divided into blocks up to `max_insert_block_size = 1048576` rows. In other words, if the `INSERT` query has less than 1048576 rows, it is made atomically.
单个数据块写入是原子的。 INSERT 的数据按每块最多 `max_insert_block_size = 1048576` 行进行分块，换句话说，如果 `INSERT` 插入的行少于 1048576，则该 INSERT 是原子的。
Data blocks are deduplicated. For multiple writes of the same data block (data blocks of the same size containing the same rows in the same order), the block is only written once. The reason for this is in case of network failures when the client application doesn't know if the data was written to the DB, so the `INSERT` query can simply be repeated. It doesn't matter which replica INSERTs were sent to with identical data. `INSERTs` are idempotent. Deduplication parameters are controlled by [merge_tree](../server_settings/settings.md) server settings.
数据块会去重。对于被多次写的相同数据块（大小相同且具有相同顺序的相同行的数据块），该块仅会写入一次。这样设计的原因是万一在网络故障时客户端应用程序不知道数据是否成功写入DB，此时可以简单地重复 `INSERT` 。把相同的数据发送给多个副本 INSERT 并不会有问题。因为这些 `INSERT` 是完全相同的（会被去重）。去重参数参看服务器设置 [merge_tree](../server_settings/settings.md) 。（注意：Replicated\*MergeTree 才会去重，不需要 zookeeper 的不带 MergeTree 不会去重）
During replication, only the source data to insert is transferred over the network. Further data transformation (merging) is coordinated and performed on all the replicas in the same way. This minimizes network usage, which means that replication works well when replicas reside in different datacenters. (Note that duplicating data in different datacenters is the main goal of replication.)
在复制期间，只有要插入的源数据通过网络传输。进一步的数据转换（合并）会在所有副本上以相同的方式进行处理执行。这样可以最大限度地减少网络使用，这意味着即使副本在不同的数据中心，数据同步也能工作良好。（能在不同数据中心中的同步数据是副本机制的主要目标。）
You can have any number of replicas of the same data. Yandex.Metrica uses double replication in production. Each server uses RAID-5 or RAID-6, and RAID-10 in some cases. This is a relatively reliable and convenient solution.
你可以给数据做任意多的副本。Yandex.Metrica 在生产中使用双副本。某一些情况下，给每台服务器都使用 RAID-5 或 RAID-6 和 RAID-10。是一种相对可靠和方便的解决方案。
The system monitors data synchronicity on replicas and is able to recover after a failure. Failover is automatic (for small differences in data) or semi-automatic (when data differs too much, which may indicate a configuration error).
系统会监视副本数据同步情况，并能在发生故障后恢复。故障转移是自动的（对于小的数据差异）或半自动的（当数据差异很大时，这可能意味是有配置错误）。
## Creating Replicated Tables
## 创建复制表 {#creating-replicated-tables}
The `Replicated` prefix is added to the table engine name. For example:`ReplicatedMergeTree`.
**Replicated\*MergeTree parameters**
在表引擎名称上加上 `Replicated` 前缀。例如：`ReplicatedMergeTree`。
- `zoo_path` — The path to the table in ZooKeeper.
- `replica_name` — The replica name in ZooKeeper.
**Replicated\*MergeTree 参数**
Example:
- `zoo_path` — ZooKeeper 中该表的路径。
- `replica_name` — ZooKeeper 中的该表的副本名称。
示例:
```sql
CREATE TABLE table_name
@ -90,7 +92,7 @@ ORDER BY (CounterID, EventDate, intHash32(UserID))
SAMPLE BY intHash32(UserID)
```
Example in deprecated syntax:
已弃用的建表语法示例：
```sql
CREATE TABLE table_name
@ -101,7 +103,7 @@ CREATE TABLE table_name
) ENGINE = ReplicatedMergeTree('/clickhouse/tables/{layer}-{shard}/hits', '{replica}', EventDate, intHash32(UserID), (CounterID, EventDate, intHash32(UserID), EventTime), 8192)
```
As the example shows, these parameters can contain substitutions in curly brackets. The substituted values are taken from the 'macros' section of the configuration file. Example:
如上例所示，这些参数可以包含宏替换的占位符，即大括号的部分。它们会被替换为配置文件里 'macros' 那部分配置的值。示例：
```xml
<macros>
@ -111,92 +113,93 @@ As the example shows, these parameters can contain substitutions in curly bracke
</macros>
```
The path to the table in ZooKeeper should be unique for each replicated table. Tables on different shards should have different paths.
In this case, the path consists of the following parts:
“ZooKeeper 中该表的路径”对每个可复制表都要是唯一的。不同分片上的表要有不同的路径。
这种情况下，路径包含下面这些部分：
`/clickhouse/tables/` is the common prefix. We recommend using exactly this one.
`/clickhouse/tables/` 是公共前缀，我们推荐使用这个。
`{layer}-{shard}` is the shard identifier. In this example it consists of two parts, since the Yandex.Metrica cluster uses bi-level sharding. For most tasks, you can leave just the {shard} substitution, which will be expanded to the shard identifier.
`{layer}-{shard}` 是分片标识部分。在此示例中，由于 Yandex.Metrica 集群使用了两级分片，所以它是由两部分组成的。但对于大多数情况来说，你只需保留 {shard} 占位符即可，它会替换展开为分片标识。
`hits` is the name of the node for the table in ZooKeeper. It is a good idea to make it the same as the table name. It is defined explicitly, because in contrast to the table name, it doesn't change after a RENAME query.
`hits` 是该表在 ZooKeeper 中的名称。使其与 ClickHouse 中的表名相同比较好。 这里它被明确定义，跟 ClickHouse 表名不一样，它并不会被 RENAME 语句修改。
The replica name identifies different replicas of the same table. You can use the server name for this, as in the example. The name only needs to be unique within each shard.
副本名称用于标识同一个表分片的不同副本。你可以使用服务器名称，如上例所示。同个分片中不同副本的副本名称要唯一。
You can define the parameters explicitly instead of using substitutions. This might be convenient for testing and for configuring small clusters. However, you can't use distributed DDL queries (`ON CLUSTER`) in this case.
你也可以显式指定这些参数，而不是使用宏替换。对于测试和配置小型集群这可能会很方便。但是，这种情况下，则不能使用分布式 DDL 语句（`ON CLUSTER`）。
When working with large clusters, we recommend using substitutions because they reduce the probability of error.
使用大型集群时，我们建议使用宏替换，因为它可以降低出错的可能性。
Run the `CREATE TABLE` query on each replica. This query creates a new replicated table, or adds a new replica to an existing one.
在每个副本服务器上运行 `CREATE TABLE` 查询。将创建新的复制表，或给现有表添加新副本。
If you add a new replica after the table already contains some data on other replicas, the data will be copied from the other replicas to the new one after running the query. In other words, the new replica syncs itself with the others.
如果其他副本上已包含了某些数据，在表上添加新副本，则在运行语句后，数据会从其他副本复制到新副本。换句话说，新副本会与其他副本同步。
To delete a replica, run `DROP TABLE`. However, only one replica is deleted – the one that resides on the server where you run the query.
要删除副本，使用 `DROP TABLE`。但它只删除那个 – 位于运行该语句的服务器上的副本。
## Recovery After Failures
## 故障恢复
If ZooKeeper is unavailable when a server starts, replicated tables switch to read-only mode. The system periodically attempts to connect to ZooKeeper.
如果服务器启动时 ZooKeeper 不可用，则复制表会切换为只读模式。系统会定期尝试去连接 ZooKeeper。
If ZooKeeper is unavailable during an `INSERT`, or an error occurs when interacting with ZooKeeper, an exception is thrown.
如果在 `INSERT` 期间 ZooKeeper 不可用，或者在与 ZooKeeper 交互时发生错误，则抛出异常。
After connecting to ZooKeeper, the system checks whether the set of data in the local file system matches the expected set of data (ZooKeeper stores this information). If there are minor inconsistencies, the system resolves them by syncing data with the replicas.
连接到 ZooKeeper 后，系统会检查本地文件系统中的数据集是否与预期的数据集（ ZooKeeper 存储此信息）一致。如果存在轻微的不一致，系统会通过与副本同步数据来解决。
If the system detects broken data parts (with the wrong size of files) or unrecognized parts (parts written to the file system but not recorded in ZooKeeper), it moves them to the 'detached' subdirectory (they are not deleted). Any missing parts are copied from the replicas.
如果系统检测到损坏的数据片段（文件大小错误）或无法识别的片段（写入文件系统但未记录在 ZooKeeper 中的部分），则会把它们移动到 'detached' 子目录（不会删除）。而副本中其他任何缺少的但正常数据片段都会被复制同步。
Note that ClickHouse does not perform any destructive actions such as automatically deleting a large amount of data.
注意，ClickHouse 不会执行任何破坏性操作，例如自动删除大量数据。
When the server starts (or establishes a new session with ZooKeeper), it only checks the quantity and sizes of all files. If the file sizes match but bytes have been changed somewhere in the middle, this is not detected immediately, but only when attempting to read the data for a `SELECT` query. The query throws an exception about a non-matching checksum or size of a compressed block. In this case, data parts are added to the verification queue and copied from the replicas if necessary.
当服务器启动（或与 ZooKeeper 建立新会话）时，它只检查所有文件的数量和大小。 如果文件大小一致但中间某处已有字节被修改过，不会立即被检测到，只有在尝试读取 `SELECT` 查询的数据时才会检测到。该查询会引发校验和不匹配或压缩块大小不一致的异常。这种情况下，数据片段会添加到验证队列中，并在必要时从其他副本中复制。
If the local set of data differs too much from the expected one, a safety mechanism is triggered. The server enters this in the log and refuses to launch. The reason for this is that this case may indicate a configuration error, such as if a replica on a shard was accidentally configured like a replica on a different shard. However, the thresholds for this mechanism are set fairly low, and this situation might occur during normal failure recovery. In this case, data is restored semi-automatically - by "pushing a button".
如果本地数据集与预期数据的差异太大，则会触发安全机制。服务器在日志中记录此内容并拒绝启动。这种情况很可能是配置错误，例如，一个分片上的副本意外配置为别的分片上的副本。然而，此机制的阈值设置得相当低，在正常故障恢复期间可能会出现这种情况。在这种情况下，数据恢复则是半自动模式，通过用户主动操作触发。
To start recovery, create the node `/path_to_table/replica_name/flags/force_restore_data` in ZooKeeper with any content, or run the command to restore all replicated tables:
要触发启动恢复，可在 ZooKeeper 中创建节点 `/path_to_table/replica_name/flags/force_restore_data`，节点值可以是任何内容，或运行命令来恢复所有的可复制表：
```bash
sudo -u clickhouse touch /var/lib/clickhouse/flags/force_restore_data
```
Then restart the server. On start, the server deletes these flags and starts recovery.
然后重启服务器。启动时，服务器会删除这些标志并开始恢复。
## Recovery After Complete Data Loss
## 在数据完全丢失后的恢复
If all data and metadata disappeared from one of the servers, follow these steps for recovery:
如果其中一个服务器的所有数据和元数据都消失了，请按照以下步骤进行恢复：
1. Install ClickHouse on the server. Define substitutions correctly in the config file that contains the shard identifier and replicas, if you use them.
2. If you had unreplicated tables that must be manually duplicated on the servers, copy their data from a replica (in the directory `/var/lib/clickhouse/data/db_name/table_name/`).
3. Copy table definitions located in `/var/lib/clickhouse/metadata/` from a replica. If a shard or replica identifier is defined explicitly in the table definitions, correct it so that it corresponds to this replica. (Alternatively, start the server and make all the `ATTACH TABLE` queries that should have been in the .sql files in `/var/lib/clickhouse/metadata/`.)
4. To start recovery, create the ZooKeeper node `/path_to_table/replica_name/flags/force_restore_data` with any content, or run the command to restore all replicated tables: `sudo -u clickhouse touch /var/lib/clickhouse/flags/force_restore_data`
1. 在服务器上安装 ClickHouse。在包含分片标识符和副本的配置文件中正确定义宏配置，如果有用到的话，
2. 如果服务器上有非复制表则必须手动复制，可以从副本服务器上（在 `/var/lib/clickhouse/data/db_name/table_name/` 目录中）复制它们的数据。
3. 从副本服务器上中复制位于 `/var/lib/clickhouse/metadata/` 中的表定义信息。如果在表定义信息中显式指定了分片或副本标识符，请更正它以使其对应于该副本。（另外，启动服务器，然后会在 `/var/lib/clickhouse/metadata/` 中的.sql文件中生成所有的 `ATTACH TABLE` 语句。）
4.要开始恢复，ZooKeeper 中创建节点 `/path_to_table/replica_name/flags/force_restore_data`，节点内容不限，或运行命令来恢复所有复制的表：`sudo -u clickhouse touch /var/lib/clickhouse/flags/force_restore_data`
Then start the server (restart, if it is already running). Data will be downloaded from replicas.
然后启动服务器（如果它已运行则重启）。数据会从副本中下载。
An alternative recovery option is to delete information about the lost replica from ZooKeeper (`/path_to_table/replica_name`), then create the replica again as described in "[Creating replicated tables](#creating-replicated-tables)".
另一种恢复方式是从 ZooKeeper（`/path_to_table/replica_name`）中删除有数据丢的副本的所有元信息，然后再按照“[创建可复制表](#creating-replicated-tables)”中的描述重新创建副本。
There is no restriction on network bandwidth during recovery. Keep this in mind if you are restoring many replicas at once.
恢复期间的网络带宽没有限制。特别注意这一点，尤其是要一次恢复很多副本。
## Converting from MergeTree to ReplicatedMergeTree
## MergeTree 转换为 ReplicatedMergeTree
We use the term `MergeTree` to refer to all table engines in the ` MergeTree family`, the same as for `ReplicatedMergeTree`.
我们使用 `MergeTree` 来表示 `MergeTree系列` 中的所有表引擎，`ReplicatedMergeTree` 同理。
If you had a `MergeTree` table that was manually replicated, you can convert it to a replicatable table. You might need to do this if you have already collected a large amount of data in a `MergeTree` table and now you want to enable replication.
如果你有一个手动同步的 `MergeTree` 表，您可以将其转换为可复制表。如果你已经在 `MergeTree` 表中收集了大量数据，并且现在要启用复制，则可以执行这些操作。
If the data differs on various replicas, first sync it, or delete this data on all the replicas except one.
如果各个副本上的数据不一致，则首先对其进行同步，或者除保留的一个副本外，删除其他所有副本上的数据。
Rename the existing MergeTree table, then create a `ReplicatedMergeTree` table with the old name.
Move the data from the old table to the 'detached' subdirectory inside the directory with the new table data (`/var/lib/clickhouse/data/db_name/table_name/`).
Then run `ALTER TABLE ATTACH PARTITION` on one of the replicas to add these data parts to the working set.
重命名现有的 MergeTree 表，然后使用旧名称创建 `ReplicatedMergeTree` 表。
将数据从旧表移动到新表（`/var/lib/clickhouse/data/db_name/table_name/`）目录内的 'detached' 目录中。
然后在其中一个副本上运行`ALTER TABLE ATTACH PARTITION`，将这些数据片段添加到工作集中。
## Converting from ReplicatedMergeTree to MergeTree
## ReplicatedMergeTree 转换为 MergeTree
Create a MergeTree table with a different name. Move all the data from the directory with the `ReplicatedMergeTree` table data to the new table's data directory. Then delete the `ReplicatedMergeTree` table and restart the server.
使用其他名称创建 MergeTree 表。将具有`ReplicatedMergeTree`表数据的目录中的所有数据移动到新表的数据目录中。然后删除`ReplicatedMergeTree`表并重新启动服务器。
如果你想在不启动服务器的情况下清除 `ReplicatedMergeTree` 表：
If you want to get rid of a `ReplicatedMergeTree` table without launching the server:
- Delete the corresponding `.sql` file in the metadata directory (`/var/lib/clickhouse/metadata/`).
- Delete the corresponding path in ZooKeeper (`/path_to_table/replica_name`).
- 删除元数据目录中的相应 `.sql` 文件（`/var/lib/clickhouse/metadata/`）。
- 删除 ZooKeeper 中的相应路径（`/path_to_table/replica_name`）。
After this, you can launch the server, create a `MergeTree` table, move the data to its directory, and then restart the server.
之后，你可以启动服务器，创建一个 `MergeTree` 表，将数据移动到其目录，然后重新启动服务器。
## Recovery When Metadata in The ZooKeeper Cluster is Lost or Damaged
## 当 ZooKeeper 集群中的元数据丢失或损坏时恢复方法
If the data in ZooKeeper was lost or damaged, you can save data by moving it to an unreplicated table as described above.
如果 ZooKeeper 中的数据丢失或损坏，如上所述，你可以通过将数据转移到非复制表来保存数据。
[Original article](https://clickhouse.yandex/docs/en/operations/table_engines/replication/) <!--hide-->
[来源文章](https://clickhouse.yandex/docs/en/operations/table_engines/replication/) <!--hide-->

78
docs/zh/operations/table_engines/summingmergetree.md
View File

@ -1,11 +1,11 @@
# SummingMergeTree
The engine inherits from [MergeTree](mergetree.md). The difference is that when merging data parts for `SummingMergeTree` tables ClickHouse replaces all the rows with the same primary key with one row which contains summarized values for the columns with the numeric data type. If the primary key is composed in a way that a single key value corresponds to large number of rows, this significantly reduces storage volume and speeds up data selection.
该引擎继承自 [MergeTree](mergetree.md)。区别在于，当合并 `SummingMergeTree` 表的数据片段时，ClickHouse 会把所有具有相同主键的行合并为一行，该行包含了被合并的行中具有数值数据类型的列的汇总值。如果主键的组合方式使得单个键值对应于大量的行，则可以显著的减少存储空间并加快数据查询的速度。
We recommend to use the engine together with `MergeTree`. Store complete data in `MergeTree` table, and use `SummingMergeTree` for aggregated data storing, for example, when preparing reports. Such an approach will prevent you from losing valuable data due to an incorrectly composed primary key.
我们推荐将该引擎和 `MergeTree` 一起使用。例如，在准备做报告的时候，将完整的数据存储在 `MergeTree` 表中，并且使用 `SummingMergeTree` 来存储聚合数据。这种方法可以使你避免因为使用不正确的主键组合方式而丢失有价值的数据。
## Creating a Table
## 建表
```
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
@ -20,23 +20,23 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
[SETTINGS name=value, ...]
```
For a description of request parameters, see [request description](../../query_language/create.md).
请求参数的描述，参考 [请求描述](../../query_language/create.md)。
**Parameters of SummingMergeTree**
**SummingMergeTree 的参数**
- `columns` - a tuple with the names of columns where values will be summarized. Optional parameter.
The columns must be of a numeric type and must not be in the primary key.
- `columns` - 包含了将要被汇总的列的列名的元组。可选参数。
所选的列必须是数值类型，并且不可位于主键中。
If `columns` not specified, ClickHouse summarizes the values in all columns with a numeric data type that are not in the primary key.
如果没有指定 `columns`，ClickHouse 会把所有不在主键中的数值类型的列都进行汇总。
**Query clauses**
**子句**
When creating a `SummingMergeTree` table the same [clauses](mergetree.md) are required, as when creating a `MergeTree` table.
创建 `SummingMergeTree` 表时，需要与创建 `MergeTree` 表时相同的[子句](mergetree.md)。
<details markdown="1"><summary>Deprecated Method for Creating a Table</summary>
<details markdown="1"><summary>已弃用的建表方法</summary>
!!! attention
Do not use this method in new projects and, if possible, switch the old projects to the method described above.
!!! attention "注意"
不要在新项目中使用该方法，可能的话，请将旧项目切换到上述方法。
```
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
@ -47,14 +47,14 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
) ENGINE [=] SummingMergeTree(date-column [, sampling_expression], (primary, key), index_granularity, [columns])
```
All of the parameters excepting `columns` have the same meaning as in `MergeTree`.
除 `columns` 外的所有参数都与 `MergeTree` 中的含义相同。
- `columns` — tuple with names of columns values of which will be summarized. Optional parameter. For a description, see the text above.
- `columns` — 包含将要被汇总的列的列名的元组。可选参数。有关说明，请参阅上文。
</details>
## Usage Example
## 用法示例
Consider the following table:
考虑如下的表：
```sql
CREATE TABLE summtt
@ -66,13 +66,13 @@ ENGINE = SummingMergeTree()
ORDER BY key
```
Insert data to it:
向其中插入数据：
```
:) INSERT INTO summtt Values(1,1),(1,2),(2,1)
```
ClickHouse may sum all the rows not completely ([see below](#data-processing)), so we use an aggregate function `sum` and `GROUP BY` clause in the query.
ClickHouse可能不会完整的汇总所有行（[见下文](#data-processing)）,因此我们在查询中使用了聚合函数 `sum` 和 `GROUP BY` 子句。
```sql
SELECT key, sum(value) FROM summtt GROUP BY key
@ -86,38 +86,38 @@ SELECT key, sum(value) FROM summtt GROUP BY key
```
## Data Processing
## 数据处理 {#data-processing}
When data are inserted into a table, they are saved as-is. Clickhouse merges the inserted parts of data periodically and this is when rows with the same primary key are summed and replaced with one for each resulting part of data.
当数据被插入到表中时，他们将被原样保存。ClickHouse 定期合并插入的数据片段，并在这个时候对所有具有相同主键的行中的列进行汇总，将这些行替换为包含汇总数据的一行记录。
ClickHouse can merge the data parts so that different resulting parts of data cat consist rows with the same primary key, i.e. the summation will be incomplete. Therefore (`SELECT`) an aggregate function [sum()](../../query_language/agg_functions/reference.md#agg_function-sum) and `GROUP BY` clause should be used in a query as described in the example above.
ClickHouse 会按片段合并数据，以至于不同的数据片段中会包含具有相同主键的行，即单个汇总片段将会是不完整的。因此，聚合函数 [sum()](../../query_language/agg_functions/reference.md#agg_function-sum) 和 `GROUP BY` 子句应该在（`SELECT`）查询语句中被使用，如上文中的例子所述。
### Common rules for summation
### 汇总的通用规则
The values in the columns with the numeric data type are summarized. The set of columns is defined by the parameter `columns`.
列中数值类型的值会被汇总。这些列的集合在参数 `columns` 中被定义。
If the values were 0 in all of the columns for summation, the row is deleted.
如果用于汇总的所有列中的值均为0，则该行会被删除。
If column is not in the primary key and is not summarized, an arbitrary value is selected from the existing ones.
如果列不在主键中且无法被汇总，则会在现有的值中任选一个。
The values are not summarized for columns in the primary key.
主键所在的列中的值不会被汇总。
### The Summation in the AggregateFunction Columns
### AggregateFunction 列中的汇总
For columns of [AggregateFunction type](../../data_types/nested_data_structures/aggregatefunction.md) ClickHouse behaves as [AggregatingMergeTree](aggregatingmergetree.md) engine aggregating according to the function.
对于 [AggregateFunction 类型](../../data_types/nested_data_structures/aggregatefunction.md)的列，ClickHouse 根据对应函数表现为 [AggregatingMergeTree](aggregatingmergetree.md) 引擎的聚合。
### Nested Structures
### 嵌套结构
Table can have nested data structures that are processed in a special way.
表中可以具有以特殊方式处理的嵌套数据结构。
If the name of a nested table ends with `Map` and it contains at least two columns that meet the following criteria:
如果嵌套表的名称以 `Map` 结尾，并且包含至少两个符合以下条件的列：
- the first column is numeric `(*Int*, Date, DateTime)`, let's call it `key`,
- the other columns are arithmetic `(*Int*, Float32/64)`, let's call it `(values...)`,
- 第一列是数值类型 `(*Int*, Date, DateTime)`，我们称之为 `key`,
- 其他的列是可计算的 `(*Int*, Float32/64)`，我们称之为 `(values...)`,
then this nested table is interpreted as a mapping of `key => (values...)`, and when merging its rows, the elements of two data sets are merged by `key` with a summation of the corresponding `(values...)`.
然后这个嵌套表会被解释为一个 `key => (values...)` 的映射，当合并它们的行时，两个数据集中的元素会被根据 `key` 合并为相应的 `(values...)` 的汇总值。
Examples:
示例：
```
[(1, 100)] + [(2, 150)] -> [(1, 100), (2, 150)]
@ -126,8 +126,8 @@ Examples:
[(1, 100), (2, 150)] + [(1, -100)] -> [(2, 150)]
```
When requesting data, use the [sumMap(key, value)](../../query_language/agg_functions/reference.md) function for aggregation of `Map`.
请求数据时，使用 [sumMap(key, value)](../../query_language/agg_functions/reference.md) 函数来对 `Map` 进行聚合。
For nested data structure, you do not need to specify its columns in the tuple of columns for summation.
对于嵌套数据结构，你无需在列的元组中指定列以进行汇总。
[Original article](https://clickhouse.yandex/docs/en/operations/table_engines/summingmergetree/) <!--hide-->
[来源文章](https://clickhouse.yandex/docs/en/operations/table_engines/summingmergetree/) <!--hide-->

6
libs/libcommon/CMakeLists.txt
View File

@ -91,9 +91,13 @@ target_include_directories (common BEFORE PRIVATE ${CCTZ_INCLUDE_DIR})
target_include_directories (common PUBLIC ${COMMON_INCLUDE_DIR})
if (NOT USE_INTERNAL_BOOST_LIBRARY)
target_include_directories (common BEFORE PUBLIC ${Boost_INCLUDE_DIRS})
target_include_directories (common SYSTEM BEFORE PUBLIC ${Boost_INCLUDE_DIRS})
endif ()
if(NOT USE_INTERNAL_POCO_LIBRARY)
target_include_directories (common SYSTEM BEFORE PUBLIC ${Poco_Foundation_INCLUDE_DIR})
endif()
target_link_libraries (common
PUBLIC
${Poco_Foundation_LIBRARY}

2
libs/libdaemon/CMakeLists.txt
View File

@ -22,4 +22,4 @@ endif ()
target_include_directories (daemon PUBLIC include)
target_link_libraries (daemon PRIVATE clickhouse_common_io clickhouse_common_config common ${Poco_Net_LIBRARY} ${Poco_Util_LIBRARY} ${EXECINFO_LIBRARY} ${ELF_LIBRARY})
target_link_libraries (daemon PRIVATE clickhouse_common_io clickhouse_common_config common ${Poco_Net_LIBRARY} ${Poco_Util_LIBRARY} ${EXECINFO_LIBRARIES})

22
libs/libmysqlxx/cmake/find_mysqlclient.cmake
View File

@ -1,17 +1,17 @@
option (ENABLE_MYSQL "Enable MySQL" ${OS_LINUX})
if (OS_LINUX)
option (USE_INTERNAL_MYSQL_LIBRARY "Set to FALSE to use system mysqlclient library instead of bundled" ${NOT_UNBUNDLED})
else ()
option (USE_INTERNAL_MYSQL_LIBRARY "Set to FALSE to use system mysqlclient library instead of bundled" OFF)
endif ()
option(ENABLE_MYSQL "Enable MySQL" 1)
if(ENABLE_MYSQL)
if(OS_LINUX)
option(USE_INTERNAL_MYSQL_LIBRARY "Set to FALSE to use system mysqlclient library instead of bundled" ${NOT_UNBUNDLED})
else()
option(USE_INTERNAL_MYSQL_LIBRARY "Set to FALSE to use system mysqlclient library instead of bundled" OFF)
endif()
if (USE_INTERNAL_MYSQL_LIBRARY AND NOT EXISTS "${ClickHouse_SOURCE_DIR}/contrib/mariadb-connector-c/README.md")
message (WARNING "submodule contrib/mariadb-connector-c is missing. to fix try run: \n git submodule update --init --recursive")
set (USE_INTERNAL_MYSQL_LIBRARY 0)
endif ()
if(USE_INTERNAL_MYSQL_LIBRARY AND NOT EXISTS "${ClickHouse_SOURCE_DIR}/contrib/mariadb-connector-c/README.md")
message(WARNING "submodule contrib/mariadb-connector-c is missing. to fix try run: \n git submodule update --init --recursive")
set(USE_INTERNAL_MYSQL_LIBRARY 0)
endif()
if (ENABLE_MYSQL)
if (USE_INTERNAL_MYSQL_LIBRARY)
set (MYSQLCLIENT_LIBRARIES mysqlclient)
set (USE_MYSQL 1)

6
libs/libmysqlxx/include/mysqlxx/Row.h
View File

@ -80,9 +80,9 @@ public:
operator private_bool_type() const { return row == nullptr ? nullptr : &Row::row; }
private:
MYSQL_ROW row = nullptr;
ResultBase * res = nullptr;
MYSQL_LENGTHS lengths;
MYSQL_ROW row{};
ResultBase * res{};
MYSQL_LENGTHS lengths{};
};
}

3
libs/libmysqlxx/src/Query.cpp
View File

@ -40,6 +40,9 @@ Query::Query(const Query & other) : std::ostream(0), conn(other.conn)
Query & Query::operator= (const Query & other)
{
if (this == &other)
return *this;
conn = other.conn;
seekp(0);

39
utils/kafka/consume.py Executable file
View File

@ -0,0 +1,39 @@
#! /usr/bin/env python3
# -*- coding: utf-8 -*-
# `pip install …`
import kafka # … kafka-python
import argparse
from pprint import pprint
def main():
parser = argparse.ArgumentParser(description='Kafka Producer client')
parser.add_argument('--server', type=str, metavar='HOST', default='localhost',
help='Kafka bootstrap-server address')
parser.add_argument('--port', type=int, metavar='PORT', default=9092,
help='Kafka bootstrap-server port')
parser.add_argument('--client', type=str, default='ch-kafka-python',
help='custom client id for this producer')
parser.add_argument('--topic', type=str, required=True,
help='name of Kafka topic to store in')
parser.add_argument('--group', type=str, required=True,
help='name of the consumer group')
args = parser.parse_args()
config = {
'bootstrap_servers': f'{args.server}:{args.port}',
'client_id': args.client,
'group_id': args.group,
}
client = kafka.KafkaConsumer(**config)
client.subscribe([args.topic])
pprint(client.poll(10000))
client.unsubscribe()
client.close()
if __name__ == "__main__":
exit(main())

41
utils/kafka/manage.py Executable file
View File

@ -0,0 +1,41 @@
#! /usr/bin/env python3
# -*- coding: utf-8 -*-
# `pip install …`
import kafka # … kafka-python
import argparse
def main():
parser = argparse.ArgumentParser(description='Kafka Topic manager')
parser.add_argument('--server', type=str, metavar='HOST', default='localhost',
help='Kafka bootstrap-server address')
parser.add_argument('--port', type=int, metavar='PORT', default=9092,
help='Kafka bootstrap-server port')
parser.add_argument('--client', type=str, default='ch-kafka-python',
help='custom client id for this producer')
commands = parser.add_mutually_exclusive_group()
commands.add_argument('--create', type=str, metavar='TOPIC', nargs='+',
help='create new topic(s) in the cluster')
commands.add_argument('--delete', type=str, metavar='TOPIC', nargs='+',
help='delete existing topic(s) from the cluster')
args = parser.parse_args()
config = {
'bootstrap_servers': f'{args.server}:{args.port}',
'client_id': args.client,
}
client = kafka.KafkaAdminClient(**config)
if args.create:
print(client.create_topics(args.create))
elif args.delete:
print(client.delete_topics(args.delete))
client.close()
if __name__ == "__main__":
exit(main())

72
utils/kafka/produce.py Executable file
View File

@ -0,0 +1,72 @@
#! /usr/bin/env python3
# -*- coding: utf-8 -*-
# `pip install …`
import kafka # … kafka-python
import argparse
from concurrent.futures import ThreadPoolExecutor
import enum
import multiprocessing
import sys
class Sync(enum.Enum):
NONE = 'none'
LEAD = 'leader'
ALL = 'all'
def __str__(self):
return self.value
def convert(self):
values = {
str(Sync.NONE): '0',
str(Sync.LEAD): '1',
str(Sync.ALL): 'all',
}
return values[self.value]
def main():
parser = argparse.ArgumentParser(description='Produce a single message taken from input')
parser.add_argument('--server', type=str, metavar='HOST', default='localhost',
help='Kafka bootstrap-server address')
parser.add_argument('--port', type=int, metavar='PORT', default=9092,
help='Kafka bootstrap-server port')
parser.add_argument('--client', type=str, default='ch-kafka-python',
help='custom client id for this producer')
parser.add_argument('--topic', type=str, required=True,
help='name of Kafka topic to store in')
parser.add_argument('--retries', type=int, default=0,
help='number of retries to send on failure')
parser.add_argument('--multiply', type=int, default=1,
help='multiplies incoming string many times')
parser.add_argument('--repeat', type=int, default=1,
help='send same (multiplied) message many times')
args = parser.parse_args()
config = {
'bootstrap_servers': f'{args.server}:{args.port}',
'client_id': args.client,
'retries': args.retries,
}
client = kafka.KafkaProducer(**config)
message = sys.stdin.buffer.read() * args.multiply
def send(num):
client.send(topic=args.topic, value=message)
print(f'iteration {num}: sent a message multiplied {args.multiply} times')
pool = ThreadPoolExecutor(max_workers=multiprocessing.cpu_count())
for num in range(args.repeat):
pool.submit(send, num)
pool.shutdown()
client.flush()
client.close()
if __name__ == "__main__":
exit(main())

52
utils/kafka/status.py Executable file
View File

@ -0,0 +1,52 @@
#! /usr/bin/env python3
# -*- coding: utf-8 -*-
# `pip install …`
import kafka # … kafka-python
import argparse
from pprint import pprint
def main():
parser = argparse.ArgumentParser(description='Kafka client to get groups and topics status')
parser.add_argument('--server', type=str, metavar='HOST', default='localhost',
help='Kafka bootstrap-server address')
parser.add_argument('--port', type=int, metavar='PORT', default=9092,
help='Kafka bootstrap-server port')
parser.add_argument('--client', type=str, default='ch-kafka-python',
help='custom client id for this producer')
args = parser.parse_args()
config = {
'bootstrap_servers': f'{args.server}:{args.port}',
'client_id': args.client,
}
client = kafka.KafkaAdminClient(**config)
consumer = kafka.KafkaConsumer(**config)
cluster = client._client.cluster
topics = cluster.topics()
for topic in topics:
print(f'Topic "{topic}":', end='')
for partition in cluster.partitions_for_topic(topic):
tp = kafka.TopicPartition(topic, partition)
print(f' {partition} (begin: {consumer.beginning_offsets([tp])[tp]}, end: {consumer.end_offsets([tp])[tp]})', end='')
print()
groups = client.list_consumer_groups()
for group in groups:
print(f'Group "{group[0]}" ({group[1]}):')
consumer = kafka.KafkaConsumer(**config, group_id=group[0])
offsets = client.list_consumer_group_offsets(group[0])
for topic, offset in offsets.items():
print(f'\t{topic.topic}[{topic.partition}]: {consumer.beginning_offsets([topic])[topic]}, {offset.offset}, {consumer.end_offsets([topic])[topic]}')
consumer.close()
client.close()
if __name__ == "__main__":
exit(main())

2
utils/s3tools/s3uploader
View File

@ -124,7 +124,7 @@ if __name__ == "__main__":
args.dataset_name, 'partitions', os.path.basename(file_path))
elif args.s3_path is not None:
s3_path = os.path.join(
args.dataset_name, s3_path, os.path.base_name(file_path))
args.dataset_name, args.s3_path, os.path.basename(file_path))
else:
raise Exception("Don't know s3-path to upload")

6
utils/test-data-generator/CMakeLists.txt
View File

@ -1,9 +1,3 @@
add_executable (test-data-generator main.cpp)
target_link_libraries(test-data-generator PRIVATE clickhouse_common_io ${Boost_PROGRAM_OPTIONS_LIBRARY})
add_executable (markov-model markov-model.cpp)
target_link_libraries(markov-model PRIVATE clickhouse_common_io ${Boost_PROGRAM_OPTIONS_LIBRARY})
if(USE_PROTOBUF)
protobuf_generate_cpp(ProtobufDelimitedMessagesSerializer_Srcs ProtobufDelimitedMessagesSerializer_Hdrs ${CMAKE_CURRENT_SOURCE_DIR}/../../dbms/tests/queries/0_stateless/00825_protobuf_format.proto)
protobuf_generate_cpp(ProtobufDelimitedMessagesSerializer_Srcs2 ProtobufDelimitedMessagesSerializer_Hdrs2 ${CMAKE_CURRENT_SOURCE_DIR}/../../dbms/tests/queries/0_stateless/00825_protobuf_format_syntax2.proto)

378
utils/test-data-generator/main.cpp
View File

@ -1,378 +0,0 @@
#include <iostream>
#include <random>
#include <pcg_random.hpp>
#include <cmath>
#include <IO/WriteBufferFromFileDescriptor.h>
#include <IO/WriteHelpers.h>
#include <IO/ReadBufferFromFile.h>
#include <Common/HashTable/Hash.h>
#include "MarkovModel.h"
/** Generate infinite stream of pseudo-random data
* like 'hits' table in Yandex.Metrica (with structure as it was in 2013)
* and output it in TabSeparated format.
*/
using namespace DB;
struct Models
{
MarkovModel Title;
MarkovModel URL;
MarkovModel SearchPhrase;
static void read(MarkovModel & model, const String & path)
{
ReadBufferFromFile in(path);
model.read(in);
}
Models()
{
read(Title, "Title.model");
read(URL, "URL.model");
read(SearchPhrase, "SearchPhrase.model");
}
};
struct Generator
{
WriteBufferFromFileDescriptor out;
pcg64 random;
pcg64 random_with_seed;
Models models;
// UInt64 WatchID = random();
String Title;
String URL;
/* String Referer;
String FlashMinor2;
String UserAgentMinor;
String MobilePhoneModel;
String Params;*/
String SearchPhrase;
/* String PageCharset;
String OriginalURL;
String BrowserLanguage;
String BrowserCountry;
String SocialNetwork;
String SocialAction;
String SocialSourcePage;
String ParamCurrency;
String OpenstatServiceName;
String OpenstatCampaignID;
String OpenstatAdID;
String OpenstatSourceID;
String UTMSource;
String UTMMedium;
String UTMCampaign;
String UTMContent;
String UTMTerm;
String FromTag;*/
Generator() : out(STDOUT_FILENO) {}
/** Choosing of distributions parameters sometimes resembles real data, but quite arbitary.
*/
void generateRow()
{
// auto gen_random64 = [&]{ return random(); };
/// Unique identifier of event.
/* WatchID += std::uniform_int_distribution<UInt64>(0, 10000000000)(random);
writeText(WatchID, out);
writeChar('\t', out);
bool JavaEnable = std::bernoulli_distribution(0.6)(random);
writeText(JavaEnable, out);
writeChar('\t', out);*/
LocalDateTime EventTime;
EventTime.year(2013);
EventTime.month(7);
EventTime.day(std::discrete_distribution<>({
0, 0, 13, 30, 0, 14, 42, 5, 6, 31, 17, 0, 0, 0, 0, 23, 10, 0, 0, 0, 19, 24, 8, 7, 0, 0, 8, 2, 15, 12, 7, 29})(random));
EventTime.hour(std::discrete_distribution<>({
13, 7, 4, 3, 2, 3, 4, 6, 10, 16, 20, 23, 24, 23, 18, 19, 19, 19, 14, 15, 14, 13, 17, 17})(random));
EventTime.minute(std::uniform_int_distribution<UInt8>(0, 59)(random));
EventTime.second(std::uniform_int_distribution<UInt8>(0, 59)(random));
UInt64 UserID = hash(4, powerLaw(5000, 1.1));
UserID = UserID / 10000000000ULL * 10000000000ULL + static_cast<time_t>(EventTime) + UserID % 1000000;
random_with_seed.seed(powerLaw(5000, 1.1));
auto get_random_with_seed = [&]{ return random_with_seed(); };
Title.resize(10000);
Title.resize(models.Title.generate(&Title[0], Title.size(), get_random_with_seed));
writeText(Title, out);
writeChar('\t', out);
/* bool GoodEvent = 1;
writeText(GoodEvent, out);
writeChar('\t', out);*/
writeText(EventTime, out);
writeChar('\t', out);
LocalDate EventDate = EventTime.toDate();
writeText(EventDate, out);
writeChar('\t', out);
UInt32 CounterID = hash(1, powerLaw(20, 1.1)) % 10000000;
writeText(CounterID, out);
writeChar('\t', out);
/* UInt32 ClientIP = hash(2, powerLaw(5000, 1.1));
writeText(ClientIP, out);
writeChar('\t', out);
UInt32 RegionID = hash(3, powerLaw(15, 1.1)) % 5000;
writeText(RegionID, out);
writeChar('\t', out);
*/
writeText(UserID, out);
writeChar('\t', out);
/* bool CounterClass = (hash(5, CounterID) % 100) < 25;
writeText(CounterClass, out);
writeChar('\t', out);
UInt8 OS = hash(6, powerLaw(10, 4)) % 100;
writeText(OS, out);
writeChar('\t', out);
UInt8 UserAgent = hash(7, powerLaw(10, 4)) % 100;
writeText(UserAgent, out);
writeChar('\t', out);
*/
URL.resize(10000);
URL.resize(models.URL.generate(&URL[0], URL.size(), get_random_with_seed));
writeText(URL, out);
writeChar('\t', out);
/// Referer
/* bool Refresh = std::bernoulli_distribution(0.1)(random);
writeText(Refresh, out);
writeChar('\t', out);
UInt16 RefererCategoryID = std::bernoulli_distribution(0.1)(random) ? 0 : (hash(8, powerLaw(10, 4)) % 10000);
writeText(RefererCategoryID, out);
writeChar('\t', out);
UInt32 RefererRegionID = std::bernoulli_distribution(0.1)(random) ? 0 : (hash(9, powerLaw(15, 1.1)) % 5000);
writeText(RefererRegionID, out);
writeChar('\t', out);
UInt16 URLCategoryID = std::bernoulli_distribution(0.1)(random) ? 0 : (hash(10, powerLaw(10, 4)) % 10000);
writeText(URLCategoryID, out);
writeChar('\t', out);
UInt32 URLRegionID = std::bernoulli_distribution(0.1)(random) ? 0 : (hash(11, powerLaw(15, 1.1)) % 5000);
writeText(URLRegionID, out);
writeChar('\t', out);
UInt16 ResolutionWidth;
UInt16 ResolutionHeight;
std::tie(ResolutionWidth, ResolutionHeight) = powerLawSampleFrom<std::pair<UInt16, UInt16>>(15, 1.1,
{
{1366, 768}, {1280, 1024}, {1920, 1080}, {0, 0}, {1024, 768},
{1280, 800}, {1440, 900}, {1600, 900}, {1600, 900}, {1680, 1050},
{768, 1024}, {1024, 600}, {1360, 768}, {1280, 720}, {1152, 864},
{1280, 768}, {320, 480}, {1920, 1200}, {320, 568}, {1093, 614},
});
if (std::bernoulli_distribution(0.1)(random))
{
ResolutionWidth = std::bernoulli_distribution(0.1)(random)
? std::uniform_int_distribution<UInt16>(160, 3000)(random)
: (std::uniform_int_distribution<UInt16>(160, 3000)(random) / 16 * 16);
ResolutionHeight = std::bernoulli_distribution(0.1)(random)
? std::uniform_int_distribution<UInt16>(160, 3000)(random)
: (ResolutionWidth / 16 * 10);
}
writeText(ResolutionWidth, out);
writeChar('\t', out);
writeText(ResolutionHeight, out);
writeChar('\t', out);
UInt8 ResolutionDepth = weightedSelect<UInt8>({32, 24, 0, 16, 8}, {2000000, 700000, 300000, 50000, 100});
writeText(ResolutionDepth, out);
writeChar('\t', out);
UInt8 FlashMajor = weightedSelect<UInt8>({11, 0, 10, 6, 9, 8, 7, 5, 12}, {2000000, 600000, 200000, 100000, 8000, 800, 600, 20, 10});
UInt8 FlashMinor = weightedSelect<UInt8>({7, 0, 8, 1, 6, 3, 2, 5, 4}, {1500000, 700000, 500000, 150000, 100000, 80000, 60000, 50000, 50000});
writeText(FlashMajor, out);
writeChar('\t', out);
writeText(FlashMinor, out);
writeChar('\t', out);
FlashMinor2.clear();
if (FlashMajor && FlashMinor)
{
FlashMinor2 = toString(hash(12, powerLaw(10, 4)) % 1000);
if (std::bernoulli_distribution(0.2)(random))
{
FlashMinor2 += '.';
FlashMinor2 += toString(hash(13, powerLaw(10, 4)) % 1000);
}
}
writeText(FlashMinor2, out);
writeChar('\t', out);
UInt8 NetMajor = weightedSelect<UInt8>({0, 3, 2, 1, 4}, {3000000, 100000, 10000, 5000, 2});
UInt8 NetMinor = weightedSelect<UInt8>({0, 5, 1}, {3000000, 200000, 5000});
writeText(NetMajor, out);
writeChar('\t', out);
writeText(NetMinor, out);
writeChar('\t', out);
UInt16 UserAgentMajor = UserAgent ? hash(14, powerLaw(10, 4)) % 100 : 0;
writeText(UserAgentMajor, out);
writeChar('\t', out);
UserAgentMinor.clear();
if (UserAgentMajor)
{
UserAgentMinor = toString(hash(15, powerLaw(10, 4)) % 100);
if (UserAgentMinor.size() == 1 && std::bernoulli_distribution(0.1)(random))
UserAgentMinor += 'a' + std::uniform_int_distribution<UInt8>(0, 25)(random);
}
writeText(UserAgentMinor, out);
writeChar('\t', out);
bool CookieEnable = std::bernoulli_distribution(0.999)(random);
writeText(CookieEnable, out);
writeChar('\t', out);
bool JavascriptEnable = std::bernoulli_distribution(0.95)(random);
writeText(JavascriptEnable, out);
writeChar('\t', out);
bool IsMobile = std::bernoulli_distribution(0.15)(random);
writeText(IsMobile, out);
writeChar('\t', out);
UInt8 MobilePhone = IsMobile ? hash(16, powerLaw(10, 4)) % 100 : 0;
writeText(MobilePhone, out);
writeChar('\t', out);
*/
/* MobilePhoneModel.resize(100);
MobilePhoneModel.resize(models.MobilePhoneModel.generate(&MobilePhoneModel[0], MobilePhoneModel.size(), gen_random64));
writeText(MobilePhoneModel, out);
writeChar('\t', out);
Params.resize(10000);
Params.resize(models.Params.generate(&Params[0], Params.size(), gen_random64));
writeText(Params, out);
writeChar('\t', out);
UInt32 IPNetworkID = hash(17, powerLaw(15, 1.1)) % 5000;
writeText(IPNetworkID, out);
writeChar('\t', out);
Int8 TraficSourceID = weightedSelect<Int8>(
{-1, 0, 1, 2, 3, 4, 5, 6, 7, 8}, {2000000, 300000, 200000, 600000, 50000, 700, 30000, 40000, 500, 2500});
writeText(TraficSourceID, out);
writeChar('\t', out);
UInt16 SearchEngineID = TraficSourceID == 2
? hash(18, powerLaw(10, 4)) % 100
: (TraficSourceID == 3
? (std::bernoulli_distribution(0.5)(random)
? hash(19, powerLaw(10, 4)) % 10
: 0)
: 0);
if (!SearchEngineID)
SearchPhrase.clear();
else
{*/
SearchPhrase.resize(1000);
SearchPhrase.resize(models.SearchPhrase.generate(&SearchPhrase[0], SearchPhrase.size(), get_random_with_seed));
// }
writeText(SearchPhrase, out);
/* writeChar('\t', out);
UInt8 AdvEngineID = weightedSelect<UInt8>(
{0, 2, 12, 17, 18, 27, 34, 36}, {3000000, 30000, 3000, 30000, 1, 100, 40, 30});
writeText(AdvEngineID, out);
writeChar('\t', out);
bool IsArtificial = std::bernoulli_distribution(0.07)(random);
writeText(IsArtificial, out);
writeChar('\t', out);
UInt16 WindowClientWidth = std::max(3000, ResolutionWidth - hash(20, UserID) % 100);
UInt16 WindowClientHeight = std::max(3000, ResolutionHeight - hash(21, UserID) % 100);
writeText(WindowClientWidth, out);
writeChar('\t', out);
writeText(WindowClientHeight, out);*/
writeChar('\n', out);
}
UInt64 powerLawImpl(double scale, double alpha, double unit_random_value)
{
return scale * std::pow(unit_random_value, -1.0 / alpha) - scale;
}
UInt64 powerLaw(double scale, double alpha)
{
return powerLawImpl(scale, alpha, std::uniform_real_distribution<double>(0, 1)(random));
}
template <typename T>
T powerLawSampleFrom(double scale, double alpha, std::initializer_list<T> set)
{
return set.begin()[powerLaw(scale, alpha) % set.size()];
}
template <typename T>
T weightedSelect(std::initializer_list<T> items, std::initializer_list<double> weights)
{
return items.begin()[std::discrete_distribution<>(weights)(random)];
}
static UInt64 hash(unsigned seed, UInt64 x)
{
return intHash64(x + seed * 0xDEADBEEF);
}
};
int main(int argc, char ** argv)
try
{
Generator generator;
while (true)
generator.generateRow();
return 0;
}
catch (...)
{
/// Broken pipe, when piped to 'head', by example.
if (errno != EPIPE)
{
std::cerr << getCurrentExceptionMessage(true) << '\n';
throw;
}
}

114
utils/test-data-generator/markov-model.cpp
View File

@ -1,114 +0,0 @@
#include <iostream>
#include <random>
#include <pcg_random.hpp>
#include <boost/program_options.hpp>
#include <IO/ReadBufferFromFileDescriptor.h>
#include <IO/WriteBufferFromFileDescriptor.h>
#include <IO/WriteBufferFromFile.h>
#include "MarkovModel.h"
using namespace DB;
int main(int argc, char ** argv)
try
{
boost::program_options::options_description desc("Allowed options");
desc.add_options()
("help,h", "produce help message")
("create", "create model")
("order", boost::program_options::value<unsigned>(), "order of model to create")
("noise", boost::program_options::value<double>(), "relative random noise to apply to created model")
("generate", "generate random strings with model")
("max-string-size", boost::program_options::value<UInt64>()->default_value(10000), "maximum size of generated string")
("limit", boost::program_options::value<UInt64>(), "stop after specified count of generated strings")
("seed", boost::program_options::value<UInt64>(), "seed passed to random number generator")
;
boost::program_options::variables_map options;
boost::program_options::store(boost::program_options::parse_command_line(argc, argv, desc), options);
auto show_usage = [&]
{
std::cout << "Usage: \n"
<< argv[0] << " --create --order=N < strings.tsv > model\n"
<< argv[0] << " --generate < model > strings.tsv\n\n";
std::cout << desc << std::endl;
};
if (options.count("help"))
{
show_usage();
return 1;
}
ReadBufferFromFileDescriptor in(STDIN_FILENO);
WriteBufferFromFileDescriptor out(STDOUT_FILENO);
pcg64 random;
if (options.count("seed"))
random.seed(options["seed"].as<UInt64>());
if (options.count("create"))
{
MarkovModel model(options["order"].as<unsigned>());
String s;
while (!in.eof())
{
readText(s, in);
assertChar('\n', in);
model.consume(s.data(), s.size());
}
if (options.count("noise"))
{
double noise = options["noise"].as<double>();
model.modifyCounts([&](UInt32 count)
{
double modified = std::normal_distribution<double>(count, count * noise)(random);
if (modified < 1)
modified = 1;
return std::round(modified);
});
}
model.write(out);
}
else if (options.count("generate"))
{
MarkovModel model;
model.read(in);
String s;
UInt64 limit = options.count("limit") ? options["limit"].as<UInt64>() : 0;
UInt64 max_string_size = options["max-string-size"].as<UInt64>();
for (size_t i = 0; limit == 0 || i < limit; ++i)
{
s.resize(max_string_size);
s.resize(model.generate(&s[0], s.size(), [&]{ return random(); }));
writeText(s, out);
writeChar('\n', out);
}
}
else
{
show_usage();
return 1;
}
return 0;
}
catch (...)
{
std::cerr << getCurrentExceptionMessage(true) << '\n';
throw;
}

Some files were not shown because too many files have changed in this diff Show More