thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-21 15:12:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #24049 from ClickHouse/aku/perf-unstable

Browse Source 
complain about unstable perf test queries
This commit is contained in:
Alexander Kuzmenkov 2021-05-24 12:59:12 +03:00 committed by GitHub
commit f227020018
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
45 changed files with 68 additions and 58 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
docker/test/performance-comparison/perf.py
View File

@ -44,7 +44,7 @@ parser.add_argument('--port', nargs='*', default=[9000], help="Space-separated l
parser.add_argument('--runs', type=int, default=1, help='Number of query runs per server.')
parser.add_argument('--max-queries', type=int, default=None, help='Test no more than this number of queries, chosen at random.')
parser.add_argument('--queries-to-run', nargs='*', type=int, default=None, help='Space-separated list of indexes of queries to test.')
parser.add_argument('--max-query-seconds', type=int, default=10, help='For how many seconds at most a query is allowed to run. The script finishes with error if this time is exceeded.')
parser.add_argument('--max-query-seconds', type=int, default=15, help='For how many seconds at most a query is allowed to run. The script finishes with error if this time is exceeded.')
parser.add_argument('--profile-seconds', type=int, default=0, help='For how many seconds to profile a query for which the performance has changed.')
parser.add_argument('--long', action='store_true', help='Do not skip the tests tagged as long.')
parser.add_argument('--print-queries', action='store_true', help='Print test queries and exit.')
@ -273,8 +273,14 @@ for query_index in queries_to_run:
prewarm_id = f'{query_prefix}.prewarm0'
try:
# Will also detect too long queries during warmup stage
res = c.execute(q, query_id = prewarm_id, settings = {'max_execution_time': args.max_query_seconds})
# During the warmup runs, we will also:
# * detect queries that are exceedingly long, to fail fast,
# * collect profiler traces, which might be helpful for analyzing
# test coverage. We disable profiler for normal runs because
# it makes the results unstable.
res = c.execute(q, query_id = prewarm_id,
settings = {'max_execution_time': args.max_query_seconds,
'query_profiler_real_time_period_ns': 10000000})
except clickhouse_driver.errors.Error as e:
# Add query id to the exception to make debugging easier.
e.args = (prewarm_id, *e.args)
@ -359,10 +365,11 @@ for query_index in queries_to_run:
# For very short queries we have a special mode where we run them for at
# least some time. The recommended lower bound of run time for "normal"
# queries is about 0.1 s, and we run them about 10 times, giving the
# time per query per server of about one second. Use this value as a
# reference for "short" queries.
# time per query per server of about one second. Run "short" queries
# for longer time, because they have a high percentage of overhead and
# might give less stable results.
if is_short[query_index]:
if server_seconds >= 2 * len(this_query_connections):
if server_seconds >= 8 * len(this_query_connections):
break
# Also limit the number of runs, so that we don't go crazy processing
# the results -- 'eqmed.sql' is really suboptimal.

14
docker/test/performance-comparison/report.py
View File

@ -446,6 +446,9 @@ if args.report == 'main':
attrs[3] = f'style="background: {color_bad}"'
else:
attrs[3] = ''
# Just don't add the slightly unstable queries we don't consider
# errors. It's not clear what the user should do with them.
continue
text += tableRow(r, attrs, anchor)
@ -553,12 +556,11 @@ if args.report == 'main':
error_tests += unstable_partial_queries
status = 'failure'
if unstable_queries:
message_array.append(str(unstable_queries) + ' unstable')
# Disabled before fix.
# if very_unstable_queries:
# status = 'failure'
# Don't show mildly unstable queries, only the very unstable ones we
# treat as errors.
if very_unstable_queries:
status = 'failure'
message_array.append(str(very_unstable_queries) + ' unstable')
error_tests += slow_average_tests
if error_tests:

2
tests/performance/agg_functions_min_max_any.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.3">
<preconditions>
<table_exists>hits_100m_single</table_exists>
</preconditions>

2
tests/performance/arithmetic.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test max_ignored_relative_change="0.7">
<settings>
<max_memory_usage>30000000000</max_memory_usage>
</settings>

2
tests/performance/array_join.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>

2
tests/performance/avg_weighted.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.4">
<test>
<settings>
<allow_experimental_bigint_types>1</allow_experimental_bigint_types>
<max_threads>1</max_threads>

2
tests/performance/bounding_ratio.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<query>SELECT boundingRatio(number, number) FROM numbers(100000000)</query>
<query>SELECT (argMax(number, number) - argMin(number, number)) / (max(number) - min(number)) FROM numbers(100000000)</query>
</test>

3
tests/performance/codecs_float_insert.xml
View File

@ -1,5 +1,4 @@
<!-- FIXME this instability is abysmal, investigate the unstable queries -->
<test max_ignored_relative_change="0.2">
<test>
<settings>
<allow_suspicious_codecs>1</allow_suspicious_codecs>
</settings>

2
tests/performance/codecs_int_insert.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<settings>
<allow_suspicious_codecs>1</allow_suspicious_codecs>
</settings>

2
tests/performance/collations.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>

2
tests/performance/date_time_64.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<preconditions>
<table_exists>hits_100m_single</table_exists>
</preconditions>

2
tests/performance/date_time_long.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test>
<substitutions>
<substitution>
<name>datetime_transform</name>

2
tests/performance/decimal_aggregates.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.5">
<test max_ignored_relative_change="0.2">
<settings>
<max_memory_usage>35G</max_memory_usage>
</settings>

2
tests/performance/decimal_casts.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.2">
<settings>
<max_memory_usage>15G</max_memory_usage>
</settings>

2
tests/performance/direct_dictionary.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test max_ignored_relative_change="0.4">
<create_query>
CREATE TABLE simple_key_direct_dictionary_source_table
(

4
tests/performance/encodeXMLComponent.xml
View File

@ -1,7 +1,7 @@
<test>
<preconditions>
<table_exists>test.hits</table_exists>
<table_exists>hits_100m_single</table_exists>
</preconditions>
<query>SELECT count() FROM test.hits WHERE NOT ignore(encodeXMLComponent(URL))</query>
<query>SELECT count() FROM hits_100m_single WHERE NOT ignore(encodeXMLComponent(URL))</query>
</test>

2
tests/performance/flat_dictionary.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test>
<create_query>
CREATE TABLE simple_key_flat_dictionary_source_table
(

2
tests/performance/float_parsing.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<substitutions>
<substitution>
<name>expr</name>

2
tests/performance/fuzz_bits.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test>

2
tests/performance/general_purpose_hashes.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.2">
<substitutions>
<substitution>
<name>gp_hash_func</name>

2
tests/performance/general_purpose_hashes_on_UUID.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.6">
<substitutions>
<substitution>
<name>hash_func</name>

2
tests/performance/generate_table_function.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<query>SELECT sum(NOT ignore(*)) FROM (SELECT * FROM generateRandom('ui64 UInt64, i64 Int64, ui32 UInt32, i32 Int32, ui16 UInt16, i16 Int16, ui8 UInt8, i8 Int8') LIMIT 1000000000);</query>
<query>SELECT sum(NOT ignore(*)) FROM (SELECT * FROM generateRandom('ui64 UInt64, i64 Int64, ui32 UInt32, i32 Int32, ui16 UInt16, i16 Int16, ui8 UInt8, i8 Int8', 0, 10, 10) LIMIT 1000000000);</query>
<query>SELECT sum(NOT ignore(*)) FROM (SELECT * FROM generateRandom('i Enum8(\'hello\' = 1, \'world\' = 5)', 0, 10, 10) LIMIT 1000000000);</query>

2
tests/performance/hashed_dictionary.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test max_ignored_relative_change="0.2">
<create_query>
CREATE TABLE simple_key_hashed_dictionary_source_table
(

2
tests/performance/int_parsing.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<preconditions>
<table_exists>hits_100m_single</table_exists>
<table_exists>hits_10m_single</table_exists>

2
tests/performance/jit_small_requests.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<query>
WITH
bitXor(number, 0x4CF2D2BAAE6DA887) AS x0,

2
tests/performance/json_extract_rapidjson.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>

2
tests/performance/json_extract_simdjson.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.2">
<substitutions>
<substitution>
<name>json</name>

2
tests/performance/math.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test max_ignored_relative_change="0.6">
<substitutions>
<substitution>
<name>func_slow</name>

2
tests/performance/merge_tree_many_partitions_2.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<create_query>CREATE TABLE bad_partitions (a UInt64, b UInt64, c UInt64, d UInt64, e UInt64, f UInt64, g UInt64, h UInt64, i UInt64, j UInt64, k UInt64, l UInt64, m UInt64, n UInt64, o UInt64, p UInt64, q UInt64, r UInt64, s UInt64, t UInt64, u UInt64, v UInt64, w UInt64, x UInt64, y UInt64, z UInt64) ENGINE = MergeTree PARTITION BY x ORDER BY x</create_query>
<fill_query>INSERT INTO bad_partitions (x) SELECT * FROM numbers_mt(3000)</fill_query>

2
tests/performance/number_formatting_formats.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test max_ignored_relative_change="0.3">
<substitutions>
<substitution>
<name>format</name>

2
tests/performance/optimized_select_final.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test>
<settings>
<do_not_merge_across_partitions_select_final>1</do_not_merge_across_partitions_select_final>
</settings>

2
tests/performance/parallel_mv.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<settings>
<parallel_view_processing>1</parallel_view_processing>
</settings>

2
tests/performance/point_in_polygon.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.2">
<settings>
<!--
Not sure why it's needed. Maybe it has something to do with the

2
tests/performance/polymorphic_parts_l.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<create_query>
CREATE TABLE hits_wide AS hits_10m_single ENGINE = MergeTree()
PARTITION BY toYYYYMM(EventDate)

2
tests/performance/polymorphic_parts_m.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<create_query>
CREATE TABLE hits_wide AS hits_10m_single ENGINE = MergeTree()
PARTITION BY toYYYYMM(EventDate)

2
tests/performance/polymorphic_parts_s.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<create_query>
CREATE TABLE hits_wide AS hits_10m_single ENGINE = MergeTree()
PARTITION BY toYYYYMM(EventDate)

2
tests/performance/questdb_sum_int32.xml
View File

@ -1,4 +1,4 @@
<test>
<test max_ignored_relative_change="0.2">
<settings>
<max_threads>4</max_threads>
<max_memory_usage>20G</max_memory_usage>

2
tests/performance/random_string.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<query>SELECT count() FROM zeros(100000000) WHERE NOT ignore(randomString(10))</query>
<query>SELECT count() FROM zeros(100000000) WHERE NOT ignore(randomString(100))</query>
<query>SELECT count() FROM zeros(1000000) WHERE NOT ignore(randomString(1000))</query>

2
tests/performance/redundant_functions_in_order_by.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test>
<preconditions>
<table_exists>hits_100m_single</table_exists>
</preconditions>

5
tests/performance/rewrite_sumIf.xml
View File

@ -1,4 +1,5 @@
<test>
<query>SELECT sumIf(1, 0) FROM numbers(100000000)</query>
<query>SELECT sumIf(1, 1) FROM numbers(100000000)</query>
<!-- Shouldn't have been a perf test, but an EXPLAIN one. -->
<query>SELECT sumIf(1, 0) FROM numbers(1000000000)</query>
<query>SELECT sumIf(1, 1) FROM numbers(1000000000)</query>
</test>

3
tests/performance/set_index.xml
View File

File diff suppressed because one or more lines are too long

2
tests/performance/sum_map.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test max_ignored_relative_change="0.2">
<settings>
<max_threads>1</max_threads>
</settings>

2
tests/performance/synthetic_hardware_benchmark.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test max_ignored_relative_change="0.2">
<settings>
<max_memory_usage>30000000000</max_memory_usage>
</settings>

2
tests/performance/url_hits.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.2">
<test>
<preconditions>
<table_exists>hits_100m_single</table_exists>
<table_exists>hits_10m_single</table_exists>

2
tests/performance/visit_param_extract_raw.xml
View File

@ -1,4 +1,4 @@
<test max_ignored_relative_change="0.3">
<test max_ignored_relative_change="0.2">
<substitutions>
<substitution>
<name>param</name>