ClickHouse/dbms/tests/performance
Mikhail Korotov 88e37020e0 JSONCompactEachRow and JSONCompactEachRowWithNamesAndTypes (#7841)
* Research commit

* Created Output EachRow Format

* Fixed bugs

* Created output format JSONCompactEachRowWithNamesAndTypes without totals

* Fixed bugs

* Fixed bugs

* Totals for JSONCompactEachRowWithNamesAndTypes

* Deleted needless debug

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Working commit

* Tests added

* Input Format for JSONCompactEachRow

* Fixed bugs for JSONCompactEachRow Input format

* Fixed bugs for JSONCompactEachRowRowInputFormat.cpp

* JSONCompactEachRow and JSONCompactEachRowWithNamesAndTypes united

* Created input format for both formats

* fixed bugs

* fixed bugs

* Working commit

* Working commit

* Working commit

* Working commit

* tests

* Working commit

* Final tests

* Performance tests added
2019-12-05 13:13:40 +03:00
..
agg_functions_min_max_any.xml
and_function.xml add perf test for function and 2019-12-03 16:28:07 +03:00
array_element.xml
array_fill.xml Create array_fill.xml 2019-10-31 14:30:11 +08:00
array_join.xml
base64_hits.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
base64.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
basename.xml
bounding_ratio.xml
cidr.xml
codec_double_delta.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
codec_gorilla.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
column_column_comparison.xml
columns_hashing.xml
complex_array_creation.xml
concat_hits.xml
conditional.xml
consistent_hashes.xml Updated performance test 2019-08-03 22:51:47 +03:00
constant_column_comparison.xml
constant_column_search.xml
count.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
cpu_synthetic.xml
create_benchmark_page.py
cryptographic_hashes.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
date_parsing.xml
date_time.xml More performance test for Date and DateTime 2019-10-15 15:20:54 +03:00
early_constant_folding.xml Add perf test for early constant folding 2019-11-01 22:33:03 +08:00
empty_string_deserialization.xml Added performance test to show degradation of performance in gcc-9 in more isolated way 2019-08-02 23:16:02 +03:00
empty_string_serialization.xml Added performance test to show degradation of performance in gcc-9 in more isolated way 2019-08-02 23:18:03 +03:00
entropy.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
first_significant_subdomain.xml
fixed_string16.xml
float_parsing.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
format_date_time.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
functions_coding.xml
functions_geo.xml
general_purpose_hashes.xml Update tests. 2019-08-20 16:33:25 +03:00
grear_circle_dist.xml Add test for greatCircleDist 2019-10-29 19:49:37 +03:00
great_circle_dist.xml Add performance test 2019-10-14 13:13:32 +03:00
group_array_moving_sum.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
h3.xml
if_array_num.xml
if_array_string.xml Changed repository URL 2019-09-23 19:18:19 +03:00
if_string_const.xml
if_string_hits.xml
insert_values_with_expressions.xml add performance test 2019-09-20 22:45:26 +03:00
inserts_arrays_lowcardinality.xml performance test for lowcardinality array, showing significant performance degradation comparing to plain array 2019-09-12 22:46:31 +02:00
int_parsing.xml
IPv4.xml
IPv6.xml
jit_large_requests.xml
jit_small_requests.xml
joins_in_memory_pmj.xml add simple perf test for partial merge join 2019-09-18 20:04:10 +03:00
joins_in_memory.xml
json_extract_rapidjson.xml Updated json perftest to loop. 2019-08-19 00:50:51 +03:00
json_extract_simdjson.xml Updated json perftest to loop. 2019-08-19 00:50:51 +03:00
leftpad.xml
linear_regression.xml Whitespaces 2019-08-09 02:55:08 +03:00
math.xml Added performance test 2019-08-01 03:37:47 +03:00
merge_table_streams.xml
merge_tree_huge_pk.xml Update merge_tree_huge_pk.xml 2019-09-18 23:07:23 +03:00
merge_tree_many_partitions_2.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
merge_tree_many_partitions.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
merge_tree_simple_select.xml Removed <name> from all performance tests #6179 2019-07-28 00:18:54 +03:00
modulo.xml const modulo optimization 2019-11-15 13:56:23 +08:00
ngram_distance.xml
number_formatting_formats.xml fixed heap buffer overflow in PacketPayloadWriteBuffer 2019-07-29 18:41:47 +03:00
nyc_taxi.xml
order_by_decimals.xml add perf test 2019-08-07 17:41:03 +03:00
order_by_read_in_order.xml better performance test for order by optimization 2019-08-02 19:17:18 +03:00
order_by_single_column.xml
parse_engine_file.xml JSONCompactEachRow and JSONCompactEachRowWithNamesAndTypes (#7841) 2019-12-05 13:13:40 +03:00
prewhere.xml PREWHERE test 2019-11-14 14:21:54 +08:00
range.xml add range performance tests 2019-11-05 15:07:25 +08:00
read_hits_with_aio.xml
README.md Add performance test 2019-10-14 13:13:32 +03:00
right.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
roundDown.xml
scalar.xml add perf test for subqueries with large scalars 2019-10-22 23:55:11 +08:00
select_format.xml JSONCompactEachRow and JSONCompactEachRowWithNamesAndTypes (#7841) 2019-12-05 13:13:40 +03:00
set_hits.xml
set.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
simple_join_query.xml
slices_hits.xml
sort.xml
string_join.xml String join perf test 2019-11-08 15:07:27 +08:00
string_set.xml IN (strings) perf test (#7739) 2019-11-14 23:04:11 +03:00
string_sort.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
system_numbers.xml
trim_numbers.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
trim_urls.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
trim_whitespace.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
uniq.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
url_hits.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
vectorize_aggregation_combinators.xml Vectorize arrayReduce and -Array combinator 2019-11-07 11:53:47 +08:00
visit_param_extract_raw.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00
website.xml Revert "Removed <name> from all performance tests #6179" 2019-07-28 00:17:44 +03:00

ClickHouse performance tests

This directory contains .xml-files with performance tests for clickhouse-performance-test tool.

How to write performance test

First of all you should check existing tests don't cover your case. If there are no such tests than you should write your own.

There two types of performance tests:

  • First is executed in loop, and have tag <type>loop</type> in config.
  • Second one is executed only once and have tag <type>once</type> in config.

Type once should be used only for endless queries. Even if your query really long (10 seconds+), it's better to choose loop test.

After you have choosen type, you have to specify preconditions. It contains table names. Only hits_100m_single, hits_10m_single, test.hits are available in CI.

The most important part of test is stop_conditions. For loop test you should always use min_time_not_changing_for_ms stop condition. For once test you can choose between average_speed_not_changing_for_ms and max_speed_not_changing_for_ms, but first is preferable. Also you should always specify total_time_ms metric. Endless tests will be ignored by CI.

metrics and main_metric settings are not important and can be ommited, because loop tests are always compared by min_time metric and once tests compared by max_rows_per_second.

You can use substitions, create, fill and drop queries to prepare test. You can find examples in this folder.

Take into account, that these tests will run in CI which consists of 56-cores and 512 RAM machines. Queries will be executed much faster than on local laptop.

If your test continued more than 10 minutes, please, add tag long to have an opportunity to run all tests and skip long ones.

How to run performance test

You have to run clickhouse-server and after you can start testing:

$ clickhouse-performance-test --input-file my_lovely_test1.xml --input-file my_lovely_test2.xml
$ clickhouse-performance-test --input-file /my_lovely_test_dir/