2020-01-23 17:48:26 +00:00
#!/usr/bin/python3
2020-04-02 18:44:58 +00:00
import argparse
2020-03-24 17:33:18 +00:00
import ast
2020-01-23 17:48:26 +00:00
import collections
import csv
2020-02-27 17:57:08 +00:00
import itertools
2020-06-11 21:24:56 +00:00
import json
2020-01-23 17:48:26 +00:00
import os
2020-06-29 14:09:51 +00:00
import os . path
2020-06-22 12:22:09 +00:00
import pprint
2020-01-23 17:48:26 +00:00
import sys
2020-02-26 16:18:58 +00:00
import traceback
2020-01-23 17:48:26 +00:00
2022-03-22 16:39:58 +00:00
parser = argparse . ArgumentParser ( description = " Create performance test report " )
parser . add_argument (
" --report " ,
default = " main " ,
choices = [ " main " , " all-queries " ] ,
help = " Which report to build " ,
)
2020-04-02 18:44:58 +00:00
args = parser . parse_args ( )
2020-08-05 15:46:05 +00:00
tables = [ ]
2020-08-05 16:07:37 +00:00
errors_explained = [ ]
2020-02-26 16:18:58 +00:00
report_errors = [ ]
2020-02-27 19:43:43 +00:00
error_tests = 0
slow_average_tests = 0
faster_queries = 0
slower_queries = 0
unstable_queries = 0
2020-03-24 17:33:18 +00:00
very_unstable_queries = 0
2023-01-22 17:43:27 +00:00
unstable_backward_incompatible_queries = 0
2020-02-26 16:18:58 +00:00
2020-04-29 17:26:28 +00:00
# max seconds to run one query by itself, not counting preparation
allowed_single_run_time = 2
2022-03-22 16:39:58 +00:00
color_bad = " #ffb0c0 "
color_good = " #b0d050 "
2020-05-20 02:19:19 +00:00
2020-04-02 18:44:58 +00:00
header_template = """
2020-01-23 17:48:26 +00:00
< ! DOCTYPE html >
2020-08-28 22:13:46 +00:00
< html lang = " en " >
2020-08-28 22:46:04 +00:00
< style >
2020-07-31 19:58:18 +00:00
body { {
2022-08-08 05:25:29 +00:00
font - family : " DejaVu Sans " , " Noto Sans " , Arial , sans - serif ;
2020-07-31 19:58:18 +00:00
background : #EEE;
} }
2020-01-23 17:48:26 +00:00
a { { color : #06F; text-decoration: none; }}
2020-07-31 19:58:18 +00:00
2020-01-23 17:48:26 +00:00
a : hover , a : active { { color : #F40; text-decoration: underline; }}
2020-07-31 19:58:18 +00:00
2020-05-09 15:50:28 +00:00
. main { { margin : auto ; max - width : 95 % ; } }
2020-02-27 17:57:08 +00:00
2020-07-31 19:58:18 +00:00
p . links a { {
padding : 5 px ; margin : 3 px ; background : #FFF; line-height: 2;
white - space : nowrap ;
box - shadow : 0 0 0 1 px rgba ( 0 , 0 , 0 , 0.05 ) , 0 8 px 25 px - 5 px rgba ( 0 , 0 , 0 , 0.1 ) ;
} }
. cancela , . cancela : link , . cancela : visited , . cancela : hover ,
. cancela : focus , . cancela : active { {
2020-02-10 16:34:07 +00:00
color : inherit ;
text - decoration : none ;
} }
2020-07-31 10:59:15 +00:00
2020-07-31 19:58:18 +00:00
table { {
border : none ;
border - spacing : 0 px ;
line - height : 1.5 ;
box - shadow : 0 0 0 1 px rgba ( 0 , 0 , 0 , 0.05 ) , 0 8 px 25 px - 5 px rgba ( 0 , 0 , 0 , 0.1 ) ;
text - align : left ;
} }
th , td { {
border : none ;
padding : 5 px ;
vertical - align : top ;
background - color : #FFF;
font - family : sans - serif ;
} }
th { {
border - bottom : 2 px solid black ;
} }
tr : nth - child ( odd ) td { { filter : brightness ( 90 % ) ; } }
2020-09-21 12:14:47 +00:00
. unexpected - query - duration tr : nth - child ( 2 ) ,
. unexpected - query - duration tr : nth - child ( 3 ) ,
. unexpected - query - duration tr : nth - child ( 5 ) ,
2020-07-31 19:58:18 +00:00
. all - query - times tr : nth - child ( 1 ) ,
. all - query - times tr : nth - child ( 2 ) ,
. all - query - times tr : nth - child ( 3 ) ,
. all - query - times tr : nth - child ( 4 ) ,
. all - query - times tr : nth - child ( 5 ) ,
. all - query - times tr : nth - child ( 7 ) ,
. changes - in - performance tr : nth - child ( 1 ) ,
. changes - in - performance tr : nth - child ( 2 ) ,
. changes - in - performance tr : nth - child ( 3 ) ,
. changes - in - performance tr : nth - child ( 4 ) ,
. changes - in - performance tr : nth - child ( 5 ) ,
. changes - in - performance tr : nth - child ( 7 ) ,
. unstable - queries tr : nth - child ( 1 ) ,
. unstable - queries tr : nth - child ( 2 ) ,
. unstable - queries tr : nth - child ( 3 ) ,
. unstable - queries tr : nth - child ( 4 ) ,
. unstable - queries tr : nth - child ( 6 ) ,
. test - performance - changes tr : nth - child ( 2 ) ,
. test - performance - changes tr : nth - child ( 3 ) ,
. test - performance - changes tr : nth - child ( 4 ) ,
. test - performance - changes tr : nth - child ( 5 ) ,
. test - performance - changes tr : nth - child ( 6 ) ,
. test - times tr : nth - child ( 2 ) ,
. test - times tr : nth - child ( 3 ) ,
. test - times tr : nth - child ( 4 ) ,
. test - times tr : nth - child ( 5 ) ,
. test - times tr : nth - child ( 6 ) ,
. test - times tr : nth - child ( 7 ) ,
. concurrent - benchmarks tr : nth - child ( 2 ) ,
. concurrent - benchmarks tr : nth - child ( 3 ) ,
. concurrent - benchmarks tr : nth - child ( 4 ) ,
. concurrent - benchmarks tr : nth - child ( 5 ) ,
. metric - changes tr : nth - child ( 2 ) ,
. metric - changes tr : nth - child ( 3 ) ,
. metric - changes tr : nth - child ( 4 ) ,
. metric - changes tr : nth - child ( 5 )
2020-07-31 10:59:15 +00:00
{ { text - align : right ; } }
2020-01-23 17:48:26 +00:00
< / style >
2020-02-26 16:18:58 +00:00
< title > Clickhouse performance comparison < / title >
2020-01-23 17:48:26 +00:00
< / head >
< body >
< div class = " main " >
2020-02-26 16:18:58 +00:00
< h1 > ClickHouse performance comparison < / h1 >
2020-04-02 18:44:58 +00:00
"""
2020-02-26 16:18:58 +00:00
2020-02-10 16:34:07 +00:00
table_anchor = 0
row_anchor = 0
2022-03-22 16:39:58 +00:00
2020-08-05 15:46:05 +00:00
def currentTableAnchor ( ) :
global table_anchor
2022-03-22 16:39:58 +00:00
return f " { table_anchor } "
2020-08-05 15:46:05 +00:00
def newTableAnchor ( ) :
2020-02-10 16:34:07 +00:00
global table_anchor
table_anchor + = 1
2020-08-05 15:46:05 +00:00
return currentTableAnchor ( )
2022-03-22 16:39:58 +00:00
2020-08-05 15:46:05 +00:00
def currentRowAnchor ( ) :
global row_anchor
global table_anchor
2022-03-22 16:39:58 +00:00
return f " { table_anchor } . { row_anchor } "
2020-02-10 16:34:07 +00:00
def nextRowAnchor ( ) :
2020-08-05 15:46:05 +00:00
global row_anchor
global table_anchor
2022-03-22 16:39:58 +00:00
return f " { table_anchor } . { row_anchor + 1 } "
2020-08-05 15:46:05 +00:00
def advanceRowAnchor ( ) :
2020-02-10 16:34:07 +00:00
global row_anchor
global table_anchor
row_anchor + = 1
2020-08-05 15:46:05 +00:00
return currentRowAnchor ( )
2020-02-10 16:34:07 +00:00
2020-08-07 01:25:45 +00:00
def tr ( x , anchor = None ) :
2022-03-22 16:39:58 +00:00
# return '<tr onclick="location.href=\'#{a}\'" id={a}>{x}</tr>'.format(a=a, x=str(x))
2020-08-07 01:25:45 +00:00
anchor = anchor if anchor else advanceRowAnchor ( )
2022-03-22 16:39:58 +00:00
return f " <tr id= { anchor } > { x } </tr> "
2020-01-23 17:48:26 +00:00
2022-03-22 16:39:58 +00:00
def td ( value , cell_attributes = " " ) :
return " <td {cell_attributes} > {value} </td> " . format (
cell_attributes = cell_attributes , value = value
)
2020-01-23 17:48:26 +00:00
2022-03-22 16:39:58 +00:00
def th ( value , cell_attributes = " " ) :
return " <th {cell_attributes} > {value} </th> " . format (
cell_attributes = cell_attributes , value = value
)
def tableRow ( cell_values , cell_attributes = [ ] , anchor = None ) :
2020-08-07 01:25:45 +00:00
return tr (
2022-03-22 16:39:58 +00:00
" " . join (
[
td ( v , a )
for v , a in itertools . zip_longest (
cell_values , cell_attributes , fillvalue = " "
)
if a is not None and v is not None
]
) ,
anchor ,
)
def tableHeader ( cell_values , cell_attributes = [ ] ) :
2020-09-23 20:45:39 +00:00
return tr (
2022-03-22 16:39:58 +00:00
" " . join (
[
th ( v , a )
for v , a in itertools . zip_longest (
cell_values , cell_attributes , fillvalue = " "
)
if a is not None and v is not None
]
)
)
2020-01-23 17:48:26 +00:00
2020-02-26 16:18:58 +00:00
def tableStart ( title ) :
2022-03-22 16:39:58 +00:00
cls = " - " . join ( title . lower ( ) . split ( " " ) [ : 3 ] )
2020-08-05 15:46:05 +00:00
global table_anchor
table_anchor = cls
anchor = currentTableAnchor ( )
2022-03-22 16:39:58 +00:00
help_anchor = " - " . join ( title . lower ( ) . split ( " " ) )
2020-07-31 10:59:15 +00:00
return f """
< h2 id = " {anchor} " >
< a class = " cancela " href = " # {anchor} " > { title } < / a >
2020-09-17 16:21:59 +00:00
< a class = " cancela " href = " https://github.com/ClickHouse/ClickHouse/tree/master/docker/test/performance-comparison# {help_anchor} " > < sup style = " color: #888 " > ? < / sup > < / a >
2020-07-31 10:59:15 +00:00
< / h2 >
< table class = " {cls} " >
"""
2020-02-26 16:18:58 +00:00
2022-03-22 16:39:58 +00:00
2020-02-26 16:18:58 +00:00
def tableEnd ( ) :
2022-03-22 16:39:58 +00:00
return " </table> "
2020-02-26 16:18:58 +00:00
2020-02-27 17:57:08 +00:00
def tsvRows ( n ) :
2020-02-26 16:18:58 +00:00
try :
2022-03-22 16:39:58 +00:00
with open ( n , encoding = " utf-8 " ) as fd :
2021-11-25 10:22:47 +00:00
result = [ ]
for row in csv . reader ( fd , delimiter = " \t " , quoting = csv . QUOTE_NONE ) :
new_row = [ ]
for e in row :
2021-11-25 12:37:00 +00:00
# The first one .encode('utf-8').decode('unicode-escape') decodes the escape characters from the strings.
# The second one (encode('latin1').decode('utf-8')) fixes the changes with unicode vs utf-8 chars, so
# 'Чем зÐ<C2B7> ни маеÑ<C2B5> ЬÑ<C2AC> Ñ<EFBFBD> ' is transformed back into 'Чем зАнимаешЬся'.
2022-03-22 16:39:58 +00:00
new_row . append (
e . encode ( " utf-8 " )
. decode ( " unicode-escape " )
. encode ( " latin1 " )
. decode ( " utf-8 " )
)
2021-11-25 10:22:47 +00:00
result . append ( new_row )
return result
2020-02-26 16:18:58 +00:00
except :
2022-03-22 16:39:58 +00:00
report_errors . append ( traceback . format_exception_only ( * sys . exc_info ( ) [ : 2 ] ) [ - 1 ] )
2020-02-26 16:18:58 +00:00
pass
return [ ]
2022-03-22 16:39:58 +00:00
2020-02-27 17:57:08 +00:00
def htmlRows ( n ) :
rawRows = tsvRows ( n )
2022-03-22 16:39:58 +00:00
result = " "
2020-02-26 16:18:58 +00:00
for row in rawRows :
result + = tableRow ( row )
2020-01-23 17:48:26 +00:00
return result
2022-03-22 16:39:58 +00:00
2020-08-05 15:46:05 +00:00
def addSimpleTable ( caption , columns , rows , pos = None ) :
global tables
2022-03-22 16:39:58 +00:00
text = " "
2020-02-27 17:57:08 +00:00
if not rows :
return
2020-08-05 15:46:05 +00:00
text + = tableStart ( caption )
text + = tableHeader ( columns )
2020-02-27 17:57:08 +00:00
for row in rows :
2020-08-05 15:46:05 +00:00
text + = tableRow ( row )
text + = tableEnd ( )
tables . insert ( pos if pos else len ( tables ) , text )
2020-02-27 17:57:08 +00:00
2022-03-22 16:39:58 +00:00
2020-08-05 15:46:05 +00:00
def add_tested_commits ( ) :
2020-04-30 08:36:33 +00:00
global report_errors
try :
2022-03-22 16:39:58 +00:00
addSimpleTable (
" Tested Commits " ,
[ " Old " , " New " ] ,
[
[
" <pre> {} </pre> " . format ( x )
for x in [
open ( " left-commit.txt " ) . read ( ) ,
open ( " right-commit.txt " ) . read ( ) ,
]
]
] ,
)
2020-04-30 08:36:33 +00:00
except :
# Don't fail if no commit info -- maybe it's a manual run.
2022-03-22 16:39:58 +00:00
report_errors . append ( traceback . format_exception_only ( * sys . exc_info ( ) [ : 2 ] ) [ - 1 ] )
2020-04-30 08:36:33 +00:00
pass
2022-03-22 16:39:58 +00:00
2020-08-05 15:46:05 +00:00
def add_report_errors ( ) :
global tables
2020-04-30 08:36:33 +00:00
global report_errors
# Add the errors reported by various steps of comparison script
try :
2022-03-22 16:39:58 +00:00
report_errors + = [ l . strip ( ) for l in open ( " report/errors.log " ) ]
2020-04-30 08:36:33 +00:00
except :
2022-03-22 16:39:58 +00:00
report_errors . append ( traceback . format_exception_only ( * sys . exc_info ( ) [ : 2 ] ) [ - 1 ] )
2020-04-30 08:36:33 +00:00
pass
2020-08-05 15:46:05 +00:00
if not report_errors :
return
2022-03-22 16:39:58 +00:00
text = tableStart ( " Errors while Building the Report " )
text + = tableHeader ( [ " Error " ] )
2020-08-05 15:46:05 +00:00
for x in report_errors :
text + = tableRow ( [ x ] )
text + = tableEnd ( )
# Insert after Tested Commits
tables . insert ( 1 , text )
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f ' <a href= " # { currentTableAnchor ( ) } " >There were some errors while building the report</a> '
]
)
2020-08-05 15:46:05 +00:00
def add_errors_explained ( ) :
2020-08-06 13:28:51 +00:00
if not errors_explained :
return
2020-08-07 01:25:45 +00:00
text = ' <a name= " fail1 " /> '
2022-03-22 16:39:58 +00:00
text + = tableStart ( " Error Summary " )
text + = tableHeader ( [ " Description " ] )
2020-08-06 13:28:51 +00:00
for row in errors_explained :
text + = tableRow ( row )
text + = tableEnd ( )
2020-08-05 15:46:05 +00:00
global tables
2020-08-06 13:28:51 +00:00
tables . insert ( 1 , text )
2020-08-05 15:46:05 +00:00
2020-04-30 08:36:33 +00:00
2022-03-22 16:39:58 +00:00
if args . report == " main " :
2020-10-02 16:54:07 +00:00
print ( ( header_template . format ( ) ) )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
add_tested_commits ( )
2022-03-22 16:39:58 +00:00
run_error_rows = tsvRows ( " run-errors.tsv " )
2020-05-22 02:51:26 +00:00
error_tests + = len ( run_error_rows )
2022-03-22 16:39:58 +00:00
addSimpleTable ( " Run Errors " , [ " Test " , " Error " ] , run_error_rows )
2020-08-05 15:46:05 +00:00
if run_error_rows :
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f ' <a href= " # { currentTableAnchor ( ) } " >There were some errors while running the tests</a> '
]
)
2020-05-22 02:51:26 +00:00
2023-01-22 17:43:27 +00:00
def add_backward_incompatible ( ) :
2022-03-22 16:39:58 +00:00
rows = tsvRows ( " report/partial-queries-report.tsv " )
2020-06-27 00:45:00 +00:00
if not rows :
return
2020-08-05 15:46:05 +00:00
2023-01-22 17:43:27 +00:00
global unstable_backward_incompatible_queries , slow_average_tests , tables
2023-01-22 19:01:55 +00:00
text = tableStart ( " Backward-incompatible queries " )
2022-03-22 16:39:58 +00:00
columns = [ " Median time, s " , " Relative time variance " , " Test " , " # " , " Query " ]
2020-08-05 15:46:05 +00:00
text + = tableHeader ( columns )
2022-03-22 16:39:58 +00:00
attrs = [ " " for c in columns ]
2020-06-27 00:45:00 +00:00
for row in rows :
2022-03-22 16:39:58 +00:00
anchor = f " { currentTableAnchor ( ) } . { row [ 2 ] } . { row [ 3 ] } "
2020-06-27 00:45:00 +00:00
if float ( row [ 1 ] ) > 0.10 :
attrs [ 1 ] = f ' style= " background: { color_bad } " '
2023-01-22 17:43:27 +00:00
unstable_backward_incompatible_queries + = 1
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f " <a href= \" # { anchor } \" >The query no. { row [ 3 ] } of test ' { row [ 2 ] } ' has excessive variance of run time. Keep it below 10%</a> "
]
)
2020-06-27 00:45:00 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 1 ] = " "
2020-06-27 00:45:00 +00:00
if float ( row [ 0 ] ) > allowed_single_run_time :
attrs [ 0 ] = f ' style= " background: { color_bad } " '
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f ' <a href= " # { anchor } " >The query no. { row [ 3 ] } of test \' { row [ 2 ] } \' is taking too long to run. Keep the run time below { allowed_single_run_time } seconds " </a> '
]
)
2020-06-27 00:45:00 +00:00
slow_average_tests + = 1
else :
2022-03-22 16:39:58 +00:00
attrs [ 0 ] = " "
2020-08-07 01:25:45 +00:00
text + = tableRow ( row , attrs , anchor )
2020-08-05 15:46:05 +00:00
text + = tableEnd ( )
tables . append ( text )
2020-06-27 00:45:00 +00:00
2023-01-22 17:43:27 +00:00
add_backward_incompatible ( )
2020-06-27 00:45:00 +00:00
2020-08-05 15:46:05 +00:00
def add_changes ( ) :
2022-03-22 16:39:58 +00:00
rows = tsvRows ( " report/changed-perf.tsv " )
2020-04-02 18:44:58 +00:00
if not rows :
return
2020-08-05 15:46:05 +00:00
global faster_queries , slower_queries , tables
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
text = tableStart ( " Changes in Performance " )
2020-04-02 18:44:58 +00:00
columns = [
2022-03-22 16:39:58 +00:00
" Old, s " , # 0
" New, s " , # 1
" Ratio of speedup (-) or slowdown (+) " , # 2
" Relative difference (new − old) / old " , # 3
" p < 0.01 threshold " , # 4
" " , # Failed # 5
" Test " , # 6
" # " , # 7
" Query " , # 8
]
attrs = [ " " for c in columns ]
2020-08-01 10:10:36 +00:00
attrs [ 5 ] = None
2020-09-23 20:45:39 +00:00
text + = tableHeader ( columns , attrs )
2020-04-02 18:44:58 +00:00
for row in rows :
2022-03-22 16:39:58 +00:00
anchor = f " { currentTableAnchor ( ) } . { row [ 6 ] } . { row [ 7 ] } "
2020-07-30 22:13:50 +00:00
if int ( row [ 5 ] ) :
2022-03-22 16:39:58 +00:00
if float ( row [ 3 ] ) < 0.0 :
2020-04-21 18:46:45 +00:00
faster_queries + = 1
2020-07-30 22:13:50 +00:00
attrs [ 2 ] = attrs [ 3 ] = f ' style= " background: { color_good } " '
2020-04-21 18:46:45 +00:00
else :
slower_queries + = 1
2020-07-30 22:13:50 +00:00
attrs [ 2 ] = attrs [ 3 ] = f ' style= " background: { color_bad } " '
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f " <a href= \" # { anchor } \" >The query no. { row [ 7 ] } of test ' { row [ 6 ] } ' has slowed down</a> "
]
)
2020-04-28 07:45:35 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 2 ] = attrs [ 3 ] = " "
2020-04-02 18:44:58 +00:00
2020-08-07 01:25:45 +00:00
text + = tableRow ( row , attrs , anchor )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
text + = tableEnd ( )
tables . append ( text )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
add_changes ( )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
def add_unstable_queries ( ) :
global unstable_queries , very_unstable_queries , tables
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
unstable_rows = tsvRows ( " report/unstable-queries.tsv " )
2020-04-02 18:44:58 +00:00
if not unstable_rows :
return
unstable_queries + = len ( unstable_rows )
columns = [
2022-03-22 16:39:58 +00:00
" Old, s " , # 0
" New, s " , # 1
" Relative difference (new - old)/old " , # 2
" p < 0.01 threshold " , # 3
" " , # Failed #4
" Test " , # 5
" # " , # 6
" Query " , # 7
2020-02-27 17:57:08 +00:00
]
2022-03-22 16:39:58 +00:00
attrs = [ " " for c in columns ]
2020-09-23 20:45:39 +00:00
attrs [ 4 ] = None
2020-02-27 17:57:08 +00:00
2022-03-22 16:39:58 +00:00
text = tableStart ( " Unstable Queries " )
2020-09-23 20:45:39 +00:00
text + = tableHeader ( columns , attrs )
2020-04-02 18:44:58 +00:00
for r in unstable_rows :
2022-03-22 16:39:58 +00:00
anchor = f " { currentTableAnchor ( ) } . { r [ 5 ] } . { r [ 6 ] } "
2020-04-28 07:45:35 +00:00
if int ( r [ 4 ] ) :
2020-04-02 18:44:58 +00:00
very_unstable_queries + = 1
2020-05-20 02:19:19 +00:00
attrs [ 3 ] = f ' style= " background: { color_bad } " '
2020-04-02 18:44:58 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 3 ] = " "
2021-05-13 09:16:43 +00:00
# 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
2020-04-02 18:44:58 +00:00
2020-08-07 01:25:45 +00:00
text + = tableRow ( r , attrs , anchor )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
text + = tableEnd ( )
2021-05-26 13:30:43 +00:00
# Don't add an empty table.
if very_unstable_queries :
tables . append ( text )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
add_unstable_queries ( )
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
skipped_tests_rows = tsvRows ( " analyze/skipped-tests.tsv " )
addSimpleTable ( " Skipped Tests " , [ " Test " , " Reason " ] , skipped_tests_rows )
addSimpleTable (
" Test Performance Changes " ,
[
" Test " ,
" Ratio of speedup (-) or slowdown (+) " ,
" Queries " ,
" Total not OK " ,
" Changed perf " ,
" Unstable " ,
] ,
tsvRows ( " report/test-perf-changes.tsv " ) ,
)
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
def add_test_times ( ) :
global slow_average_tests , tables
2022-03-22 16:39:58 +00:00
rows = tsvRows ( " report/test-times.tsv " )
2020-04-02 18:44:58 +00:00
if not rows :
return
columns = [
2022-03-22 16:39:58 +00:00
" Test " , # 0
" Wall clock time, entire test, s " , # 1
" Total client time for measured query runs, s " , # 2
" Queries " , # 3
" Longest query, total for measured runs, s " , # 4
" Wall clock time per query, s " , # 5
" Shortest query, total for measured runs, s " , # 6
" " , # Runs #7
]
attrs = [ " " for c in columns ]
2020-09-23 07:28:28 +00:00
attrs [ 7 ] = None
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
text = tableStart ( " Test Times " )
2020-09-23 20:45:39 +00:00
text + = tableHeader ( columns , attrs )
2020-05-09 15:58:47 +00:00
2022-03-22 16:39:58 +00:00
allowed_average_run_time = 3.75 # 60 seconds per test at (7 + 1) * 2 runs
2020-04-02 18:44:58 +00:00
for r in rows :
2022-03-22 16:39:58 +00:00
anchor = f " { currentTableAnchor ( ) } . { r [ 0 ] } "
2020-09-18 15:06:41 +00:00
total_runs = ( int ( r [ 7 ] ) + 1 ) * 2 # one prewarm run, two servers
2022-03-22 16:39:58 +00:00
if r [ 0 ] != " Total " and float ( r [ 5 ] ) > allowed_average_run_time * total_runs :
2020-04-02 18:44:58 +00:00
# FIXME should be 15s max -- investigate parallel_insert
slow_average_tests + = 1
2020-09-03 01:42:25 +00:00
attrs [ 5 ] = f ' style= " background: { color_bad } " '
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f " <a href= \" # { anchor } \" >The test ' { r [ 0 ] } ' is too slow to run as a whole. Investigate whether the create and fill queries can be sped up "
]
)
2020-04-02 18:44:58 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 5 ] = " "
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
if r [ 0 ] != " Total " and float ( r [ 4 ] ) > allowed_single_run_time * total_runs :
2020-04-02 18:44:58 +00:00
slow_average_tests + = 1
2020-09-03 01:42:25 +00:00
attrs [ 4 ] = f ' style= " background: { color_bad } " '
2022-03-22 16:39:58 +00:00
errors_explained . append (
[
f " <a href= \" ./all-queries.html#all-query-times. { r [ 0 ] } .0 \" >Some query of the test ' { r [ 0 ] } ' is too slow to run. See the all queries report "
]
)
2020-04-02 18:44:58 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 4 ] = " "
2020-04-02 18:44:58 +00:00
2020-09-01 19:05:57 +00:00
text + = tableRow ( r , attrs , anchor )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
text + = tableEnd ( )
tables . append ( text )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
add_test_times ( )
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
addSimpleTable (
" Metric Changes " ,
[
" Metric " ,
" Old median value " ,
" New median value " ,
" Relative difference " ,
" Times difference " ,
] ,
tsvRows ( " metrics/changes.tsv " ) ,
)
2020-06-25 20:19:27 +00:00
2020-08-05 15:46:05 +00:00
add_report_errors ( )
add_errors_explained ( )
for t in tables :
print ( t )
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
print (
f """
2020-08-28 22:46:04 +00:00
< / div >
2020-04-02 18:44:58 +00:00
< p class = " links " >
< a href = " all-queries.html " > All queries < / a >
< a href = " compare.log " > Log < / a >
2020-05-20 02:19:19 +00:00
< a href = " output.7z " > Test output < / a >
2021-04-23 13:47:13 +00:00
{ os . getenv ( " CHPC_ADD_REPORT_LINKS " ) or ' ' }
2020-04-02 18:44:58 +00:00
< / p >
< / body >
< / html >
2022-03-22 16:39:58 +00:00
"""
)
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
status = " success "
message = " See the report "
2020-04-02 18:44:58 +00:00
message_array = [ ]
if slow_average_tests :
2022-03-22 16:39:58 +00:00
status = " failure "
message_array . append ( str ( slow_average_tests ) + " too long " )
2020-04-02 18:44:58 +00:00
if faster_queries :
2022-03-22 16:39:58 +00:00
message_array . append ( str ( faster_queries ) + " faster " )
2020-04-02 18:44:58 +00:00
if slower_queries :
2023-05-30 18:53:45 +00:00
# This threshold should be synchronized with the value in https://github.com/ClickHouse/ClickHouse/blob/master/tests/ci/performance_comparison_check.py#L225
# False positives rate should be < 1%: https://shorturl.at/CDEK8
if slower_queries > 5 :
2022-03-22 16:39:58 +00:00
status = " failure "
message_array . append ( str ( slower_queries ) + " slower " )
2020-04-02 18:44:58 +00:00
2023-01-22 17:43:27 +00:00
if unstable_backward_incompatible_queries :
very_unstable_queries + = unstable_backward_incompatible_queries
2022-03-22 16:39:58 +00:00
status = " failure "
2020-06-27 00:45:00 +00:00
2021-05-12 22:32:53 +00:00
# Don't show mildly unstable queries, only the very unstable ones we
# treat as errors.
2021-05-12 09:47:08 +00:00
if very_unstable_queries :
2022-03-22 16:39:58 +00:00
message_array . append ( str ( very_unstable_queries ) + " unstable " )
2022-05-04 14:56:43 +00:00
# FIXME: uncomment the following lines when tests are stable and
# reliable
# if very_unstable_queries > 5:
# error_tests += very_unstable_queries
# status = "failure"
#
2022-05-04 11:02:34 +00:00
# error_tests += slow_average_tests
2022-05-04 14:56:43 +00:00
# FIXME: until here
2022-05-04 11:02:34 +00:00
2020-04-02 18:44:58 +00:00
if error_tests :
2022-03-22 16:39:58 +00:00
status = " failure "
message_array . insert ( 0 , str ( error_tests ) + " errors " )
2020-04-02 18:44:58 +00:00
if message_array :
2022-03-22 16:39:58 +00:00
message = " , " . join ( message_array )
2020-04-02 18:44:58 +00:00
if report_errors :
2022-03-22 16:39:58 +00:00
status = " failure "
message = " Errors while building the report. "
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
print (
(
"""
2020-04-02 18:44:58 +00:00
< ! - - status : { status } - - >
< ! - - message : { message } - - >
2022-03-22 16:39:58 +00:00
""" .format(
status = status , message = message
)
)
)
2020-04-02 18:44:58 +00:00
2022-03-22 16:39:58 +00:00
elif args . report == " all-queries " :
2020-10-02 16:54:07 +00:00
print ( ( header_template . format ( ) ) )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
add_tested_commits ( )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
def add_all_queries ( ) :
2022-03-22 16:39:58 +00:00
rows = tsvRows ( " report/all-queries.tsv " )
2020-04-02 18:44:58 +00:00
if not rows :
return
columns = [
2022-03-22 16:39:58 +00:00
" " , # Changed #0
" " , # Unstable #1
" Old, s " , # 2
" New, s " , # 3
" Ratio of speedup (-) or slowdown (+) " , # 4
" Relative difference (new − old) / old " , # 5
" p < 0.01 threshold " , # 6
" Test " , # 7
" # " , # 8
" Query " , # 9
]
attrs = [ " " for c in columns ]
2020-04-28 07:45:35 +00:00
attrs [ 0 ] = None
attrs [ 1 ] = None
2020-09-23 20:45:39 +00:00
2022-03-22 16:39:58 +00:00
text = tableStart ( " All Query Times " )
2020-09-23 20:45:39 +00:00
text + = tableHeader ( columns , attrs )
2020-04-02 18:44:58 +00:00
for r in rows :
2022-03-22 16:39:58 +00:00
anchor = f " { currentTableAnchor ( ) } . { r [ 7 ] } . { r [ 8 ] } "
2020-04-28 07:45:35 +00:00
if int ( r [ 1 ] ) :
2020-05-20 02:19:19 +00:00
attrs [ 6 ] = f ' style= " background: { color_bad } " '
2020-04-21 18:46:45 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 6 ] = " "
2020-04-21 18:46:45 +00:00
2020-04-28 07:45:35 +00:00
if int ( r [ 0 ] ) :
2022-03-22 16:39:58 +00:00
if float ( r [ 5 ] ) > 0.0 :
2020-07-30 22:13:50 +00:00
attrs [ 4 ] = attrs [ 5 ] = f ' style= " background: { color_bad } " '
2020-04-21 18:46:45 +00:00
else :
2020-07-30 22:13:50 +00:00
attrs [ 4 ] = attrs [ 5 ] = f ' style= " background: { color_good } " '
2020-04-02 18:44:58 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 4 ] = attrs [ 5 ] = " "
2020-04-02 18:44:58 +00:00
2020-04-29 17:26:28 +00:00
if ( float ( r [ 2 ] ) + float ( r [ 3 ] ) ) / 2 > allowed_single_run_time :
2020-05-20 02:19:19 +00:00
attrs [ 2 ] = f ' style= " background: { color_bad } " '
attrs [ 3 ] = f ' style= " background: { color_bad } " '
2020-04-29 17:26:28 +00:00
else :
2022-03-22 16:39:58 +00:00
attrs [ 2 ] = " "
attrs [ 3 ] = " "
2020-04-29 17:26:28 +00:00
2020-08-07 01:25:45 +00:00
text + = tableRow ( r , attrs , anchor )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
text + = tableEnd ( )
tables . append ( text )
2020-04-02 18:44:58 +00:00
2020-08-05 15:46:05 +00:00
add_all_queries ( )
add_report_errors ( )
for t in tables :
print ( t )
2020-04-30 08:36:33 +00:00
2022-03-22 16:39:58 +00:00
print (
f """
2020-08-28 22:46:04 +00:00
< / div >
2020-04-02 18:44:58 +00:00
< p class = " links " >
< a href = " report.html " > Main report < / a >
< a href = " compare.log " > Log < / a >
2020-05-20 02:19:19 +00:00
< a href = " output.7z " > Test output < / a >
2021-04-23 13:47:13 +00:00
{ os . getenv ( " CHPC_ADD_REPORT_LINKS " ) or ' ' }
2020-04-02 18:44:58 +00:00
< / p >
< / body >
< / html >
2022-03-22 16:39:58 +00:00
"""
)