ClickHouse performance comparison
""".format())
table_template = """
"""
table_anchor = 0
row_anchor = 0
def nextTableAnchor():
global table_anchor
table_anchor += 1
return str(table_anchor)
def nextRowAnchor():
global row_anchor
global table_anchor
row_anchor += 1
return str(table_anchor) + "." + str(row_anchor)
def tr(x):
a = nextRowAnchor()
#return '
{x}
'.format(a=a, x=str(x))
return '
{x}
'.format(a=a, x=str(x))
def td(value, cell_class = ''):
return '
{value} | '.format(
cell_class = cell_class,
value = value)
def th(x):
return '
' + str(x) + ' | '
def tableRow(r):
return tr(''.join([td(f) for f in r]))
def tableHeader(r):
return tr(''.join([th(f) for f in r]))
def tableStart(title):
return """
""".format(
anchor = nextTableAnchor(),
title = title)
def tableEnd():
return '
'
def tsvRowsRaw(n):
result = []
try:
with open(n, encoding='utf-8') as fd:
return [row for row in csv.reader(fd, delimiter="\t", quotechar='"')]
except:
report_errors.append(
traceback.format_exception_only(
*sys.exc_info()[:2])[-1])
pass
return []
def tsvTableRows(n):
rawRows = tsvRowsRaw(n)
result = ''
for row in rawRows:
result += tableRow(row)
return result
print(table_template.format(
anchor = nextTableAnchor(),
caption = 'Tested commits',
header = tableHeader(['Old', 'New']),
rows = tableRow([open('left-commit.txt').read(), open('right-commit.txt').read()])))
print(table_template.format(
anchor = nextTableAnchor(),
caption = 'Changes in performance',
header = tableHeader(['Old, s', 'New, s',
'Relative difference (new - old)/old',
'Randomization distribution quantiles [5%, 50%, 95%]',
'Test', 'Query']),
rows = tsvTableRows('changed-perf.tsv')))
print(table_template.format(
anchor = nextTableAnchor(),
caption = 'Slow on client',
header = tableHeader(['Client time, s', 'Server time, s', 'Ratio', 'Query']),
rows = tsvTableRows('slow-on-client.tsv')))
print(table_template.format(
anchor = nextTableAnchor(),
caption = 'Unstable queries',
header = tableHeader(['Old, s', 'New, s',
'Relative difference (new - old)/old',
'Randomization distribution quantiles [5%, 50%, 95%]',
'Test', 'Query']),
rows = tsvTableRows('unstable-queries.tsv')))
print(table_template.format(
anchor = nextTableAnchor(),
caption = 'Run errors',
header = tableHeader(['A', 'B']),
rows = tsvTableRows('run-errors.log')))
print(table_template.format(
anchor = nextTableAnchor(),
caption = 'Tests with most unstable queries',
header = tableHeader(['Test', 'Unstable', 'Changed perf', 'Total not OK']),
rows = tsvTableRows('bad-tests.tsv')))
def print_test_times():
rows = tsvRowsRaw('test-times.tsv')
if not rows:
return
print(rows, file=sys.stderr)
print(tableStart('Test times'))
print(tableHeader([
'Test', #0
'Wall clock time, s', #1
'Total client time, s', #2
'Total queries', #3
'Ignored short queries', #4
'Longest query
(sum for all runs), s', #5
'Avg wall clock time
(sum for all runs), s', #6
'Shortest query
(sum for all runs), s', #7
]))
for r in rows:
print(tableRow(r))
print(tableEnd())
print_test_times()
if len(report_errors):
print(tableStart('Errors while building the report'))
print(tableHeader(['Error']))
for x in report_errors:
print(tableRow([x]))
print(tableEnd())
print("""
Test output