ClickHouse/docker/test/performance-comparison/compare.sh

1302 lines
50 KiB
Bash
Raw Normal View History

#!/bin/bash
2020-08-19 15:34:23 +00:00
set -exu
set -o pipefail
trap "exit" INT TERM
2021-03-15 18:45:57 +00:00
# The watchdog is in the separate process group, so we have to kill it separately
# if the script terminates earlier.
trap 'kill $(jobs -pr) ${watchdog_pid:-} ||:' EXIT
2020-02-27 20:02:50 +00:00
stage=${stage:-}
2020-03-24 17:33:18 +00:00
script_dir="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
# upstream/master
LEFT_SERVER_PORT=9001
# patched version
RIGHT_SERVER_PORT=9002
2020-08-19 15:08:23 +00:00
function wait_for_server # port, pid
{
for _ in {1..60}
do
if clickhouse-client --port "$1" --query "select 1" || ! kill -0 "$2"
then
break
fi
sleep 1
done
if ! clickhouse-client --port "$1" --query "select 1"
then
echo "Cannot connect to ClickHouse server at $1"
return 1
fi
if ! kill -0 "$2"
then
echo "Server pid '$2' is not running"
return 1
fi
}
function left_or_right()
{
local from=$1 && shift
local basename=$1 && shift
if [ -e "$from/$basename" ]; then
echo "$from/$basename"
return
fi
case "$from" in
left) echo "right/$basename" ;;
right) echo "left/$basename" ;;
esac
}
function configure
{
2020-02-25 19:51:09 +00:00
# Use the new config for both servers, so that we can change it in a PR.
2019-12-26 21:33:10 +00:00
rm right/config/config.d/text_log.xml ||:
2020-02-25 19:51:09 +00:00
cp -rv right/config left ||:
2020-01-14 19:05:58 +00:00
# Start a temporary server to rename the tables
2020-04-17 15:47:01 +00:00
while killall clickhouse-server; do echo . ; sleep 1 ; done
2020-01-14 19:05:58 +00:00
echo all killed
set -m # Spawn temporary in its own process groups
local setup_left_server_opts=(
# server options
--config-file=left/config/config.xml
--
# server *config* directives overrides
--path db0
--user_files_path db0/user_files
--top_level_domains_path "$(left_or_right right top_level_domains)"
--tcp_port $LEFT_SERVER_PORT
)
left/clickhouse-server "${setup_left_server_opts[@]}" &> setup-server-log.log &
2020-01-14 19:05:58 +00:00
left_pid=$!
kill -0 $left_pid
disown $left_pid
set +m
2020-08-19 15:08:23 +00:00
wait_for_server $LEFT_SERVER_PORT $left_pid
2020-08-19 15:08:23 +00:00
echo Server for setup started
2020-01-14 19:05:58 +00:00
clickhouse-client --port $LEFT_SERVER_PORT --query "create database test" ||:
clickhouse-client --port $LEFT_SERVER_PORT --query "rename table datasets.hits_v1 to test.hits" ||:
2020-02-14 12:55:47 +00:00
2020-04-17 15:47:01 +00:00
while killall clickhouse-server; do echo . ; sleep 1 ; done
2020-02-14 12:55:47 +00:00
echo all killed
2020-01-14 19:05:58 +00:00
# Make copies of the original db for both servers. Use hardlinks instead
2020-06-25 20:19:27 +00:00
# of copying to save space. Before that, remove preprocessed configs and
# system tables, because sharing them between servers with hardlinks may
2020-10-27 11:04:03 +00:00
# lead to weird effects.
2020-01-14 19:05:58 +00:00
rm -r left/db ||:
rm -r right/db ||:
2020-04-13 19:06:07 +00:00
rm -r db0/preprocessed_configs ||:
2020-06-25 20:19:27 +00:00
rm -r db0/{data,metadata}/system ||:
rm db0/status ||:
2020-01-14 19:05:58 +00:00
cp -al db0/ left/db/
cp -al db0/ right/db/
2020-02-25 19:51:09 +00:00
}
function restart
{
2020-04-17 15:47:01 +00:00
while killall clickhouse-server; do echo . ; sleep 1 ; done
2020-02-25 19:51:09 +00:00
echo all killed
2020-01-14 19:05:58 +00:00
2020-10-26 16:35:37 +00:00
# Change the jemalloc settings here.
2020-09-25 10:19:37 +00:00
# https://github.com/jemalloc/jemalloc/wiki/Getting-Started
2020-10-26 16:35:37 +00:00
export MALLOC_CONF="confirm_conf:true"
2020-09-25 10:19:37 +00:00
set -m # Spawn servers in their own process groups
2020-10-27 11:04:03 +00:00
local left_server_opts=(
# server options
--config-file=left/config/config.xml
--
# server *config* directives overrides
--path left/db
--user_files_path left/db/user_files
--top_level_domains_path "$(left_or_right left top_level_domains)"
--tcp_port $LEFT_SERVER_PORT
)
left/clickhouse-server "${left_server_opts[@]}" &>> left-server-log.log &
left_pid=$!
kill -0 $left_pid
disown $left_pid
local right_server_opts=(
# server options
--config-file=right/config/config.xml
--
# server *config* directives overrides
--path right/db
--user_files_path right/db/user_files
--top_level_domains_path "$(left_or_right right top_level_domains)"
--tcp_port $RIGHT_SERVER_PORT
)
right/clickhouse-server "${right_server_opts[@]}" &>> right-server-log.log &
right_pid=$!
kill -0 $right_pid
disown $right_pid
set +m
2020-09-25 10:19:37 +00:00
unset MALLOC_CONF
wait_for_server $LEFT_SERVER_PORT $left_pid
echo left ok
2020-08-19 15:08:23 +00:00
wait_for_server $RIGHT_SERVER_PORT $right_pid
echo right ok
2019-12-26 21:33:10 +00:00
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.tables where database != 'system'"
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.build_options"
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.tables where database != 'system'"
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.build_options"
2020-04-17 15:47:01 +00:00
# Check again that both servers we started are running -- this is important
# for running locally, when there might be some other servers started and we
# will connect to them instead.
kill -0 $left_pid
kill -0 $right_pid
}
function run_tests
{
# Just check that the script runs at all
"$script_dir/perf.py" --help > /dev/null
2020-04-30 08:36:33 +00:00
# Find the directory with test files.
if [ -v CHPC_TEST_PATH ]
2020-02-14 12:55:47 +00:00
then
2020-04-30 08:36:33 +00:00
# Use the explicitly set path to directory with test files.
test_prefix="$CHPC_TEST_PATH"
elif [ "$PR_TO_TEST" == "0" ]
2020-04-30 08:36:33 +00:00
then
# When testing commits from master, use the older test files. This
# allows the tests to pass even when we add new functions and tests for
# them, that are not supported in the old revision.
test_prefix=left/performance
2020-06-05 14:30:26 +00:00
else
2020-04-30 08:36:33 +00:00
# For PRs, use newer test files so we can test these changes.
test_prefix=right/performance
2020-02-14 12:55:47 +00:00
fi
2020-04-30 08:36:33 +00:00
# Determine which tests to run.
2020-04-23 20:18:46 +00:00
if [ -v CHPC_TEST_GREP ]
2020-02-14 12:55:47 +00:00
then
2020-04-30 08:36:33 +00:00
# Run only explicitly specified tests, if any.
2020-08-19 15:31:13 +00:00
# shellcheck disable=SC2010
2020-04-24 03:23:38 +00:00
test_files=$(ls "$test_prefix" | grep "$CHPC_TEST_GREP" | xargs -I{} -n1 readlink -f "$test_prefix/{}")
elif [ "$PR_TO_TEST" -ne 0 ] \
&& [ "$(wc -l < changed-test-definitions.txt)" -gt 0 ] \
&& [ "$(wc -l < changed-test-scripts.txt)" -eq 0 ] \
&& [ "$(wc -l < other-changed-files.txt)" -eq 0 ]
2020-04-02 18:44:58 +00:00
then
# If only the perf tests were changed in the PR, we will run only these
# tests. The lists of changed files are prepared in entrypoint.sh because
# it has the repository.
test_files=$(sed "s/tests\/performance/${test_prefix//\//\\/}/" changed-test-definitions.txt)
2020-04-30 08:36:33 +00:00
else
# The default -- run all tests found in the test dir.
2020-04-02 18:44:58 +00:00
test_files=$(ls "$test_prefix"/*.xml)
fi
# For PRs w/o changes in test definitons and scripts, test only a subset of
# queries, and run them less times. If the corresponding environment variables
# are already set, keep those values.
if [ "$PR_TO_TEST" -ne 0 ] \
&& [ "$(wc -l < changed-test-definitions.txt)" -eq 0 ] \
2020-09-21 12:01:27 +00:00
&& [ "$(wc -l < changed-test-scripts.txt)" -eq 0 ]
then
2020-09-16 09:06:42 +00:00
CHPC_RUNS=${CHPC_RUNS:-7}
2020-09-16 09:16:07 +00:00
CHPC_MAX_QUERIES=${CHPC_MAX_QUERIES:-20}
2020-09-21 12:01:27 +00:00
else
CHPC_RUNS=${CHPC_RUNS:-13}
CHPC_MAX_QUERIES=${CHPC_MAX_QUERIES:-0}
fi
2020-09-15 10:44:21 +00:00
export CHPC_RUNS
export CHPC_MAX_QUERIES
2020-06-23 12:30:45 +00:00
# Determine which concurrent benchmarks to run. For now, the only test
# we run as a concurrent benchmark is 'website'. Run it as benchmark if we
# are also going to run it as a normal test.
2020-08-19 15:31:13 +00:00
for test in $test_files; do echo "$test"; done | sed -n '/website/p' > benchmarks-to-run.txt
2020-06-23 12:30:45 +00:00
2020-04-30 08:36:33 +00:00
# Delete old report files.
for x in {test-times,wall-clock-times}.tsv
2020-04-30 08:36:33 +00:00
do
rm -v "$x" ||:
touch "$x"
done
2020-06-25 20:19:27 +00:00
# Randomize test order.
test_files=$(for f in $test_files; do echo "$f"; done | sort -R)
2020-09-25 09:52:09 +00:00
# Limit profiling time to 10 minutes, not to run for too long.
profile_seconds_left=600
2020-02-25 19:51:09 +00:00
# Run the tests.
2021-04-23 17:32:30 +00:00
total_tests=$(echo "$test_files" | wc -w)
2021-04-23 13:47:13 +00:00
current_test=0
2020-02-25 19:51:09 +00:00
test_name="<none>"
2020-02-14 12:55:47 +00:00
for test in $test_files
do
2021-04-23 19:01:24 +00:00
echo "$current_test of $total_tests tests complete" > status.txt
2020-06-25 20:19:27 +00:00
# Check that both servers are alive, and restart them if they die.
clickhouse-client --port $LEFT_SERVER_PORT --query "select 1 format Null" \
2020-06-25 20:19:27 +00:00
|| { echo $test_name >> left-server-died.log ; restart ; }
clickhouse-client --port $RIGHT_SERVER_PORT --query "select 1 format Null" \
2020-06-25 20:19:27 +00:00
|| { echo $test_name >> right-server-died.log ; restart ; }
2020-02-25 19:51:09 +00:00
2020-02-27 20:02:50 +00:00
test_name=$(basename "$test" ".xml")
echo test "$test_name"
2020-02-14 19:11:46 +00:00
2020-09-25 09:52:09 +00:00
# Don't profile if we're past the time limit.
2020-10-27 11:04:03 +00:00
# Use awk because bash doesn't support floating point arithmetic.
2020-09-29 11:07:31 +00:00
profile_seconds=$(awk "BEGIN { print ($profile_seconds_left > 0 ? 10 : 0) }")
2020-09-25 09:52:09 +00:00
2020-02-06 12:46:57 +00:00
TIMEFORMAT=$(printf "$test_name\t%%3R\t%%3U\t%%3S\n")
2020-09-14 17:25:17 +00:00
# The grep is to filter out set -x output and keep only time output.
# The '2>&1 >/dev/null' redirects stderr to stdout, and discards stdout.
2020-06-29 08:29:40 +00:00
{ \
time "$script_dir/perf.py" --host localhost localhost --port $LEFT_SERVER_PORT $RIGHT_SERVER_PORT \
--runs "$CHPC_RUNS" --max-queries "$CHPC_MAX_QUERIES" \
2020-09-25 09:52:09 +00:00
--profile-seconds "$profile_seconds" \
2020-06-29 08:29:40 +00:00
-- "$test" > "$test_name-raw.tsv" 2> "$test_name-err.log" ; \
2020-09-14 17:25:17 +00:00
} 2>&1 >/dev/null | tee >(grep -v ^+ >> "wall-clock-times.tsv") \
2020-06-29 08:29:40 +00:00
|| echo "Test $test_name failed with error code $?" >> "$test_name-err.log"
2020-09-25 09:52:09 +00:00
profile_seconds_left=$(awk -F' ' \
2020-09-28 11:44:49 +00:00
'BEGIN { s = '$profile_seconds_left'; } /^profile-total/ { s -= $2 } END { print s }' \
2020-09-25 09:52:09 +00:00
"$test_name-raw.tsv")
2021-04-23 13:47:13 +00:00
current_test=$((current_test + 1))
done
2020-02-14 12:55:47 +00:00
2020-02-14 19:11:46 +00:00
unset TIMEFORMAT
2020-02-14 12:55:47 +00:00
wait
2020-02-14 19:11:46 +00:00
}
2020-02-14 12:55:47 +00:00
2020-04-21 21:53:13 +00:00
function get_profiles_watchdog
{
sleep 600
2020-04-21 21:53:13 +00:00
2020-04-22 19:41:40 +00:00
echo "The trace collection did not finish in time." >> profile-errors.log
2020-04-21 21:53:13 +00:00
for pid in $(pgrep -f clickhouse)
do
2020-04-29 09:47:17 +00:00
gdb -p "$pid" --batch --ex "info proc all" --ex "thread apply all bt" --ex quit &> "$pid.gdb.log" &
2020-04-21 21:53:13 +00:00
done
wait
2020-04-29 09:47:17 +00:00
for _ in {1..10}
2020-04-21 21:53:13 +00:00
do
if ! pkill -f clickhouse
then
break
fi
sleep 1
done
}
2020-02-14 19:11:46 +00:00
function get_profiles
{
2020-02-14 12:55:47 +00:00
# Collect the profiles
clickhouse-client --port $LEFT_SERVER_PORT --query "set query_profiler_cpu_time_period_ns = 0"
clickhouse-client --port $LEFT_SERVER_PORT --query "set query_profiler_real_time_period_ns = 0"
clickhouse-client --port $LEFT_SERVER_PORT --query "system flush logs" &
2020-06-25 20:19:27 +00:00
clickhouse-client --port $RIGHT_SERVER_PORT --query "set query_profiler_cpu_time_period_ns = 0"
clickhouse-client --port $RIGHT_SERVER_PORT --query "set query_profiler_real_time_period_ns = 0"
clickhouse-client --port $RIGHT_SERVER_PORT --query "system flush logs" &
2020-06-25 20:19:27 +00:00
wait
2020-04-17 15:47:01 +00:00
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.query_log where type = 2 format TSVWithNamesAndTypes" > left-query-log.tsv ||: &
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.query_thread_log format TSVWithNamesAndTypes" > left-query-thread-log.tsv ||: &
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.trace_log format TSVWithNamesAndTypes" > left-trace-log.tsv ||: &
clickhouse-client --port $LEFT_SERVER_PORT --query "select arrayJoin(trace) addr, concat(splitByChar('/', addressToLine(addr))[-1], '#', demangle(addressToSymbol(addr)) ) name from system.trace_log group by addr format TSVWithNamesAndTypes" > left-addresses.tsv ||: &
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.metric_log format TSVWithNamesAndTypes" > left-metric-log.tsv ||: &
clickhouse-client --port $LEFT_SERVER_PORT --query "select * from system.asynchronous_metric_log format TSVWithNamesAndTypes" > left-async-metric-log.tsv ||: &
2020-04-17 15:47:01 +00:00
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.query_log where type = 2 format TSVWithNamesAndTypes" > right-query-log.tsv ||: &
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.query_thread_log format TSVWithNamesAndTypes" > right-query-thread-log.tsv ||: &
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.trace_log format TSVWithNamesAndTypes" > right-trace-log.tsv ||: &
clickhouse-client --port $RIGHT_SERVER_PORT --query "select arrayJoin(trace) addr, concat(splitByChar('/', addressToLine(addr))[-1], '#', demangle(addressToSymbol(addr)) ) name from system.trace_log group by addr format TSVWithNamesAndTypes" > right-addresses.tsv ||: &
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.metric_log format TSVWithNamesAndTypes" > right-metric-log.tsv ||: &
clickhouse-client --port $RIGHT_SERVER_PORT --query "select * from system.asynchronous_metric_log format TSVWithNamesAndTypes" > right-async-metric-log.tsv ||: &
2020-02-14 12:55:47 +00:00
wait
2020-04-24 07:19:23 +00:00
# Just check that the servers are alive so that we return a proper exit code.
# We don't consistently check the return codes of the above background jobs.
clickhouse-client --port $LEFT_SERVER_PORT --query "select 1"
clickhouse-client --port $RIGHT_SERVER_PORT --query "select 1"
}
2020-05-21 23:03:41 +00:00
function build_log_column_definitions
2020-03-02 18:47:37 +00:00
{
# FIXME This loop builds column definitons from TSVWithNamesAndTypes in an
# absolutely atrocious way. This should be done by the file() function itself.
2020-06-16 08:21:15 +00:00
for x in {right,left}-{addresses,{query,query-thread,trace,{async-,}metric}-log}.tsv
do
paste -d' ' \
<(sed -n '1{s/\t/\n/g;p;q}' "$x" | sed 's/\(^.*$\)/"\1"/') \
<(sed -n '2{s/\t/\n/g;p;q}' "$x" ) \
| tr '\n' ', ' | sed 's/,$//' > "$x.columns"
done
2020-05-21 23:03:41 +00:00
}
# Build and analyze randomization distribution for all queries.
function analyze_queries
{
rm -v analyze-commands.txt analyze-errors.log all-queries.tsv unstable-queries.tsv ./*-report.tsv raw-queries.tsv ||:
2020-05-25 01:42:56 +00:00
rm -rf analyze ||:
mkdir analyze analyze/tmp ||:
2020-05-21 23:03:41 +00:00
build_log_column_definitions
2020-04-29 17:26:28 +00:00
# Split the raw test output into files suitable for analysis.
# To debug calculations only for a particular test, substitute a suitable
# wildcard here, e.g. `for test_file in modulo-raw.tsv`.
2020-08-19 15:31:13 +00:00
for test_file in *-raw.tsv
2020-04-29 17:26:28 +00:00
do
test_name=$(basename "$test_file" "-raw.tsv")
sed -n "s/^query\t/$test_name\t/p" < "$test_file" >> "analyze/query-runs.tsv"
2020-09-25 09:52:09 +00:00
sed -n "s/^profile\t/$test_name\t/p" < "$test_file" >> "analyze/query-profiles.tsv"
sed -n "s/^client-time\t/$test_name\t/p" < "$test_file" >> "analyze/client-times.tsv"
sed -n "s/^report-threshold\t/$test_name\t/p" < "$test_file" >> "analyze/report-thresholds.tsv"
sed -n "s/^skipped\t/$test_name\t/p" < "$test_file" >> "analyze/skipped-tests.tsv"
sed -n "s/^display-name\t/$test_name\t/p" < "$test_file" >> "analyze/query-display-names.tsv"
sed -n "s/^short\t/$test_name\t/p" < "$test_file" >> "analyze/marked-short-queries.tsv"
sed -n "s/^partial\t/$test_name\t/p" < "$test_file" >> "analyze/partial-queries.tsv"
2020-04-29 17:26:28 +00:00
done
2020-04-27 12:47:59 +00:00
# for each query run, prepare array of metrics from query log
2020-05-25 01:42:56 +00:00
clickhouse-local --query "
create view query_runs as select * from file('analyze/query-runs.tsv', TSV,
'test text, query_index int, query_id text, version UInt8, time float');
2020-07-03 08:39:43 +00:00
-- Separately process 'partial' queries which we could only run on the new server
-- because they use new functions. We can't make normal stats for them, but still
-- have to show some stats so that the PR author can tweak them.
create view partial_queries as select test, query_index
from file('analyze/partial-queries.tsv', TSV,
'test text, query_index int, servers Array(int)');
create table partial_query_times engine File(TSVWithNamesAndTypes,
'analyze/partial-query-times.tsv')
as select test, query_index, stddevPop(time) time_stddev, median(time) time_median
from query_runs
where (test, query_index) in partial_queries
group by test, query_index
;
2020-07-03 08:39:43 +00:00
-- Process queries that were run normally, on both servers.
create view left_query_log as select *
from file('left-query-log.tsv', TSVWithNamesAndTypes,
'$(cat "left-query-log.tsv.columns")');
create view right_query_log as select *
from file('right-query-log.tsv', TSVWithNamesAndTypes,
'$(cat "right-query-log.tsv.columns")');
2020-06-25 20:19:27 +00:00
create view query_logs as
select 0 version, query_id, ProfileEvents.Names, ProfileEvents.Values,
2020-08-11 11:53:25 +00:00
query_duration_ms, memory_usage from left_query_log
2020-06-25 20:19:27 +00:00
union all
select 1 version, query_id, ProfileEvents.Names, ProfileEvents.Values,
2020-08-11 11:53:25 +00:00
query_duration_ms, memory_usage from right_query_log
2020-06-25 20:19:27 +00:00
;
-- This is a single source of truth on all metrics we have for query runs. The
-- metrics include ProfileEvents from system.query_log, and query run times
-- reported by the perf.py test runner.
create table query_run_metric_arrays engine File(TSV, 'analyze/query-run-metric-arrays.tsv')
2020-06-25 20:19:27 +00:00
as
with (
-- sumMapState with the list of all keys with '-0.' values. Negative zero is because
-- sumMap removes keys with positive zeros.
with (select groupUniqArrayArray(ProfileEvents.Names) from query_logs) as all_names
select arrayReduce('sumMapState', [(all_names, arrayMap(x->-0., all_names))])
) as all_metrics
select test, query_index, version, query_id,
(finalizeAggregation(
arrayReduce('sumMapMergeState',
[
all_metrics,
arrayReduce('sumMapState',
[(ProfileEvents.Names,
arrayMap(x->toFloat64(x), ProfileEvents.Values))]
),
arrayReduce('sumMapState', [(
2020-08-11 11:53:25 +00:00
['client_time', 'server_time', 'memory_usage'],
2020-06-25 20:19:27 +00:00
arrayMap(x->if(x != 0., x, -0.), [
toFloat64(query_runs.time),
2020-08-11 11:53:25 +00:00
toFloat64(query_duration_ms / 1000.),
toFloat64(memory_usage)]))])
2020-06-25 20:19:27 +00:00
]
)) as metrics_tuple).1 metric_names,
metrics_tuple.2 metric_values
from query_logs
right join query_runs
2020-06-25 20:19:27 +00:00
on query_logs.query_id = query_runs.query_id
and query_logs.version = query_runs.version
where (test, query_index) not in partial_queries
2020-06-25 20:19:27 +00:00
;
-- This is just for convenience -- human-readable + easy to make plots.
create table query_run_metrics_denorm engine File(TSV, 'analyze/query-run-metrics-denorm.tsv')
as select test, query_index, metric_names, version, query_id, metric_values
from query_run_metric_arrays
array join metric_names, metric_values
order by test, query_index, metric_names, version, query_id
;
-- Filter out tests that don't have an even number of runs, to avoid breaking
-- the further calculations. This may happen if there was an error during the
-- test runs, e.g. the server died. It will be reported in test errors, so we
-- don't have to report it again.
create view broken_queries as
select test, query_index
from query_runs
group by test, query_index
having count(*) % 2 != 0
;
-- This is for statistical processing with eqmed.sql
create table query_run_metrics_for_stats engine File(
2020-06-25 20:19:27 +00:00
TSV, -- do not add header -- will parse with grep
'analyze/query-run-metrics-for-stats.tsv')
as select test, query_index, 0 run, version,
-- For debugging, add a filter for a particular metric like this:
-- arrayFilter(m, n -> n = 'client_time', metric_values, metric_names)
-- metric_values
-- Note that further reporting may break, because the metric names are
-- not filtered.
metric_values
from query_run_metric_arrays
where (test, query_index) not in broken_queries
2020-06-25 20:19:27 +00:00
order by test, query_index, run, version
;
-- This is the list of metric names, so that we can join them back after
-- statistical processing.
2020-06-25 20:19:27 +00:00
create table query_run_metric_names engine File(TSV, 'analyze/query-run-metric-names.tsv')
as select metric_names from query_run_metric_arrays limit 1
;
" 2> >(tee -a analyze/errors.log 1>&2)
2020-04-27 12:47:59 +00:00
# This is a lateral join in bash... please forgive me.
# We don't have arrayPermute(), so I have to make random permutations with
2020-04-27 12:47:59 +00:00
# `order by rand`, and it becomes really slow if I do it for more than one
# query. We also don't have lateral joins. So I just put all runs of each
# query into a separate file, and then compute randomization distribution
# for each file. I do this in parallel using GNU parallel.
( set +x # do not bloat the log
2020-04-27 12:47:59 +00:00
IFS=$'\n'
for prefix in $(cut -f1,2 "analyze/query-run-metrics-for-stats.tsv" | sort | uniq)
2020-04-27 12:47:59 +00:00
do
2020-08-19 15:31:13 +00:00
file="analyze/tmp/${prefix// /_}.tsv"
grep "^$prefix " "analyze/query-run-metrics-for-stats.tsv" > "$file" &
printf "%s\0\n" \
"clickhouse-local \
--file \"$file\" \
--structure 'test text, query text, run int, version UInt8, metrics Array(float)' \
--query \"$(cat "$script_dir/eqmed.sql")\" \
>> \"analyze/query-metric-stats.tsv\"" \
2>> analyze/errors.log \
>> analyze/commands.txt
2020-04-27 12:47:59 +00:00
done
wait
unset IFS
)
2020-04-27 12:47:59 +00:00
2020-10-01 10:56:56 +00:00
# The comparison script might be bound to one NUMA node for better test
# stability, and the calculation runs out of memory because of this. Use
# all nodes.
2020-10-16 14:44:59 +00:00
numactl --show
2020-10-19 14:31:02 +00:00
numactl --cpunodebind=all --membind=all numactl --show
numactl --cpunodebind=all --membind=all parallel --joblog analyze/parallel-log.txt --null < analyze/commands.txt 2>> analyze/errors.log
clickhouse-local --query "
-- Join the metric names back to the metric statistics we've calculated, and make
-- a denormalized table of them -- statistics for all metrics for all queries.
-- The WITH, ARRAY JOIN and CROSS JOIN do not like each other:
-- https://github.com/ClickHouse/ClickHouse/issues/11868
-- https://github.com/ClickHouse/ClickHouse/issues/11757
-- Because of this, we make a view with arrays first, and then apply all the
-- array joins.
create view query_metric_stat_arrays as
with (select * from file('analyze/query-run-metric-names.tsv',
TSV, 'n Array(String)')) as metric_name
select test, query_index, metric_name, left, right, diff, stat_threshold
from file('analyze/query-metric-stats.tsv', TSV, 'left Array(float),
right Array(float), diff Array(float), stat_threshold Array(float),
test text, query_index int') reports
order by test, query_index, metric_name
;
create table query_metric_stats_denorm engine File(TSVWithNamesAndTypes,
'analyze/query-metric-stats-denorm.tsv')
as select test, query_index, metric_name, left, right, diff, stat_threshold
from query_metric_stat_arrays
left array join metric_name, left, right, diff, stat_threshold
order by test, query_index, metric_name
;
" 2> >(tee -a analyze/errors.log 1>&2)
2020-03-02 18:47:37 +00:00
}
# Analyze results
2020-02-07 18:34:24 +00:00
function report
{
2020-04-28 07:45:35 +00:00
rm -r report ||:
2020-05-25 01:42:56 +00:00
mkdir report report/tmp ||:
2020-04-28 07:45:35 +00:00
2020-05-28 07:45:03 +00:00
rm ./*.{rep,svg} test-times.tsv test-dump.tsv unstable.tsv unstable-query-ids.tsv unstable-query-metrics.tsv changed-perf.tsv unstable-tests.tsv unstable-queries.tsv bad-tests.tsv slow-on-client.tsv all-queries.tsv run-errors.tsv ||:
2020-02-17 19:32:40 +00:00
2020-05-21 23:03:41 +00:00
build_log_column_definitions
cat analyze/errors.log >> report/errors.log ||:
2020-04-28 07:45:35 +00:00
cat profile-errors.log >> report/errors.log ||:
2020-04-22 19:41:40 +00:00
2020-04-17 15:47:01 +00:00
clickhouse-local --query "
create view query_display_names as select * from
file('analyze/query-display-names.tsv', TSV,
'test text, query_index int, query_display_name text')
;
2020-07-03 08:39:43 +00:00
create view partial_query_times as select * from
file('analyze/partial-query-times.tsv', TSVWithNamesAndTypes,
'test text, query_index int, time_stddev float, time_median float')
;
-- Report for partial queries that we could only run on the new server (e.g.
-- queries with new functions added in the tested PR).
create table partial_queries_report engine File(TSV, 'report/partial-queries-report.tsv')
2020-07-31 19:58:18 +00:00
as select toDecimal64(time_median, 3) time,
toDecimal64(time_stddev / time_median, 3) relative_time_stddev,
test, query_index, query_display_name
2020-07-03 08:39:43 +00:00
from partial_query_times
join query_display_names using (test, query_index)
order by test, query_index
;
create view query_metric_stats as
select * from file('analyze/query-metric-stats-denorm.tsv',
TSVWithNamesAndTypes,
'test text, query_index int, metric_name text, left float, right float,
diff float, stat_threshold float')
2020-06-25 20:19:27 +00:00
;
-- Main statistics for queries -- query time as reported in query log.
2020-04-28 07:45:35 +00:00
create table queries engine File(TSVWithNamesAndTypes, 'report/queries.tsv')
2020-03-23 09:39:12 +00:00
as select
-- It is important to have a non-strict inequality with stat_threshold
-- here. The randomization distribution is actually discrete, and when
-- the number of runs is small, the quantile we need (e.g. 0.99) turns
-- out to be the maximum value of the distribution. We can also hit this
-- maximum possible value with our test run, and this obviously means
-- that we have observed the difference to the best precision possible
-- for the given number of runs. If we use a strict equality here, we
-- will miss such cases. This happened in the wild and lead to some
-- uncaught regressions, because for the default 7 runs we do for PRs,
-- the randomization distribution has only 16 values, so the max quantile
-- is actually 0.9375.
abs(diff) > report_threshold and abs(diff) >= stat_threshold as changed_fail,
abs(diff) > report_threshold - 0.05 and abs(diff) >= stat_threshold as changed_show,
2020-10-27 11:04:03 +00:00
2020-09-03 01:42:25 +00:00
not changed_fail and stat_threshold > report_threshold + 0.10 as unstable_fail,
not changed_show and stat_threshold > report_threshold - 0.05 as unstable_show,
2020-10-27 11:04:03 +00:00
2020-04-28 07:45:35 +00:00
left, right, diff, stat_threshold,
2020-06-27 01:15:58 +00:00
if(report_threshold > 0, report_threshold, 0.10) as report_threshold,
2020-07-15 13:29:00 +00:00
query_metric_stats.test test, query_metric_stats.query_index query_index,
query_display_name
from query_metric_stats
left join file('analyze/report-thresholds.tsv', TSV,
'test text, report_threshold float') thresholds
on query_metric_stats.test = thresholds.test
2020-07-03 08:39:43 +00:00
left join query_display_names
on query_metric_stats.test = query_display_names.test
and query_metric_stats.query_index = query_display_names.query_index
2020-09-01 19:05:57 +00:00
-- 'server_time' is rounded down to ms, which might be bad for very short queries.
-- Use 'client_time' instead.
where metric_name = 'client_time'
order by test, query_index, metric_name
;
2020-04-28 07:45:35 +00:00
2020-07-03 08:39:43 +00:00
create table changed_perf_report engine File(TSV, 'report/changed-perf.tsv') as
2020-08-05 23:02:18 +00:00
with
2020-07-31 19:58:18 +00:00
-- server_time is sometimes reported as zero (if it's less than 1 ms),
-- so we have to work around this to not get an error about conversion
-- of NaN to decimal.
2020-08-05 23:02:18 +00:00
(left > right ? left / right : right / left) as times_change_float,
isFinite(times_change_float) as times_change_finite,
toDecimal64(times_change_finite ? times_change_float : 1., 3) as times_change_decimal,
times_change_finite
? (left > right ? '-' : '+') || toString(times_change_decimal) || 'x'
: '--' as times_change_str
select
toDecimal64(left, 3), toDecimal64(right, 3), times_change_str,
toDecimal64(diff, 3), toDecimal64(stat_threshold, 3),
changed_fail, test, query_index, query_display_name
from queries where changed_show order by abs(diff) desc;
2020-02-11 20:00:53 +00:00
2020-07-03 08:39:43 +00:00
create table unstable_queries_report engine File(TSV, 'report/unstable-queries.tsv') as
2020-07-31 19:58:18 +00:00
select
toDecimal64(left, 3), toDecimal64(right, 3), toDecimal64(diff, 3),
toDecimal64(stat_threshold, 3), unstable_fail, test, query_index, query_display_name
from queries where unstable_show order by stat_threshold desc;
2020-02-11 20:00:53 +00:00
2020-06-09 13:29:07 +00:00
2020-09-03 00:57:25 +00:00
create view test_speedup as
select
test,
exp2(avg(log2(left / right))) times_speedup,
count(*) queries,
unstable + changed bad,
sum(changed_show) changed,
sum(unstable_show) unstable
2020-06-09 13:29:07 +00:00
from queries
group by test
2020-09-03 00:57:25 +00:00
order by times_speedup desc
;
create view total_speedup as
select
'Total' test,
exp2(avg(log2(times_speedup))) times_speedup,
sum(queries) queries,
unstable + changed bad,
sum(changed) changed,
sum(unstable) unstable
from test_speedup
2020-06-09 13:29:07 +00:00
;
2020-07-03 08:39:43 +00:00
create table test_perf_changes_report engine File(TSV, 'report/test-perf-changes.tsv') as
2020-09-03 00:57:25 +00:00
with
(times_speedup >= 1
? '-' || toString(toDecimal64(times_speedup, 3)) || 'x'
: '+' || toString(toDecimal64(1 / times_speedup, 3)) || 'x')
as times_speedup_str
select test, times_speedup_str, queries, bad, changed, unstable
-- Not sure what's the precedence of UNION ALL vs WHERE & ORDER BY, hence all
-- the braces.
from (
(
select * from total_speedup
) union all (
select * from test_speedup
where
(times_speedup >= 1 ? times_speedup : (1 / times_speedup)) >= 1.005
or bad
)
)
2020-09-06 23:49:07 +00:00
order by test = 'Total' desc, times_speedup desc
2020-06-09 13:29:07 +00:00
;
2020-02-07 18:34:24 +00:00
2020-09-03 00:57:25 +00:00
2020-07-03 08:39:43 +00:00
create view total_client_time_per_query as select *
from file('analyze/client-times.tsv', TSV,
'test text, query_index int, client float, server float');
2020-02-07 18:34:24 +00:00
2020-07-03 08:39:43 +00:00
create table slow_on_client_report engine File(TSV, 'report/slow-on-client.tsv') as
2020-07-31 19:58:18 +00:00
select client, server, toDecimal64(client/server, 3) p,
test, query_display_name
2020-07-03 08:39:43 +00:00
from total_client_time_per_query left join query_display_names using (test, query_index)
2020-07-31 19:58:18 +00:00
where p > toDecimal64(1.02, 3) order by p desc;
2020-02-10 16:34:07 +00:00
2020-07-03 08:39:43 +00:00
create table wall_clock_time_per_test engine Memory as select *
from file('wall-clock-times.tsv', TSV, 'test text, real float, user float, system float');
2020-02-10 16:34:07 +00:00
create table test_time engine Memory as
2020-02-11 15:01:16 +00:00
select test, sum(client) total_client_time,
2020-09-03 01:42:25 +00:00
max(client) query_max,
min(client) query_min,
count(*) queries
2020-07-03 08:39:43 +00:00
from total_client_time_per_query full join queries using (test, query_index)
2020-02-11 15:01:16 +00:00
group by test;
2020-02-07 18:34:24 +00:00
create view query_runs as select * from file('analyze/query-runs.tsv', TSV,
'test text, query_index int, query_id text, version UInt8, time float');
--
-- Guess the number of query runs used for this test. The number is required to
-- calculate and check the average query run time in the report.
-- We have to be careful, because we will encounter:
-- 1) partial queries which run only on one server
-- 2) short queries which run for a much higher number of times
-- 3) some errors that make query run for a different number of times on a
-- particular server.
--
create view test_runs as
select test,
-- Default to 7 runs if there are only 'short' queries in the test, and
-- we can't determine the number of runs.
if((ceil(medianOrDefaultIf(t.runs, not short), 0) as r) != 0, r, 7) runs
from (
select
-- The query id is the same for both servers, so no need to divide here.
uniqExact(query_id) runs,
(test, query_index) in
(select * from file('analyze/marked-short-queries.tsv', TSV,
'test text, query_index int'))
as short,
test, query_index
from query_runs
group by test, query_index
) t
group by test
;
2020-09-24 11:46:03 +00:00
create view test_times_view as
select
wall_clock_time_per_test.test test,
real,
total_client_time,
2020-02-11 15:01:16 +00:00
queries,
2020-09-24 11:46:03 +00:00
query_max,
2021-03-18 14:59:49 +00:00
real / if(queries > 0, queries, 1) avg_real_per_query,
2020-09-24 11:46:03 +00:00
query_min,
runs
2020-04-29 16:27:40 +00:00
from test_time
-- wall clock times are also measured for skipped tests, so don't
-- do full join
left join wall_clock_time_per_test
on wall_clock_time_per_test.test = test_time.test
full join test_runs
on test_runs.test = test_time.test
2020-09-24 11:46:03 +00:00
;
-- WITH TOTALS doesn't work with INSERT SELECT, so we have to jump through these
-- hoops: https://github.com/ClickHouse/ClickHouse/issues/15227
create view test_times_view_total as
select
'Total' test,
sum(real),
sum(total_client_time),
sum(queries),
max(query_max),
2021-03-18 14:59:49 +00:00
sum(real) / if(sum(queries) > 0, sum(queries), 1) avg_real_per_query,
2020-09-24 11:46:03 +00:00
min(query_min),
-- Totaling the number of runs doesn't make sense, but use the max so
-- that the reporting script doesn't complain about queries being too
-- long.
max(runs)
from test_times_view
;
create table test_times_report engine File(TSV, 'report/test-times.tsv') as
select
test,
toDecimal64(real, 3),
toDecimal64(total_client_time, 3),
queries,
toDecimal64(query_max, 3),
toDecimal64(avg_real_per_query, 3),
toDecimal64(query_min, 3),
runs
from (
select * from test_times_view
union all
select * from test_times_view_total
)
order by test = 'Total' desc, avg_real_per_query desc
;
2020-02-10 16:34:07 +00:00
-- report for all queries page, only main metric
2020-07-03 08:39:43 +00:00
create table all_tests_report engine File(TSV, 'report/all-queries.tsv') as
2020-08-05 23:02:18 +00:00
with
-- server_time is sometimes reported as zero (if it's less than 1 ms),
-- so we have to work around this to not get an error about conversion
-- of NaN to decimal.
(left > right ? left / right : right / left) as times_change_float,
isFinite(times_change_float) as times_change_finite,
toDecimal64(times_change_finite ? times_change_float : 1., 3) as times_change_decimal,
times_change_finite
? (left > right ? '-' : '+') || toString(times_change_decimal) || 'x'
: '--' as times_change_str
2020-04-28 07:45:35 +00:00
select changed_fail, unstable_fail,
2020-08-05 23:02:18 +00:00
toDecimal64(left, 3), toDecimal64(right, 3), times_change_str,
2020-07-31 19:58:18 +00:00
toDecimal64(isFinite(diff) ? diff : 0, 3),
toDecimal64(isFinite(stat_threshold) ? stat_threshold : 0, 3),
test, query_index, query_display_name
2020-06-08 13:57:33 +00:00
from queries order by test, query_index;
2020-04-28 07:45:35 +00:00
2020-09-25 09:52:09 +00:00
-- Report of queries that have inconsistent 'short' markings:
-- 1) have short duration, but are not marked as 'short'
-- 2) the reverse -- marked 'short' but take too long.
-- The threshold for 2) is significantly larger than the threshold for 1), to
-- avoid jitter.
create view shortness
2020-10-27 11:04:03 +00:00
as select
(test, query_index) in
(select * from file('analyze/marked-short-queries.tsv', TSV,
'test text, query_index int'))
as marked_short,
time, test, query_index, query_display_name
2020-07-03 08:39:43 +00:00
from (
2020-09-03 01:42:25 +00:00
select right time, test, query_index from queries
2020-07-03 08:39:43 +00:00
union all
select time_median, test, query_index from partial_query_times
) times
left join query_display_names
on times.test = query_display_names.test
and times.query_index = query_display_names.query_index
;
create table inconsistent_short_marking_report
engine File(TSV, 'report/unexpected-query-duration.tsv')
as select
2020-09-23 08:21:55 +00:00
multiIf(marked_short and time > 0.1, '\"short\" queries must run faster than 0.02 s',
not marked_short and time < 0.02, '\"normal\" queries must run longer than 0.1 s',
'') problem,
marked_short, time,
test, query_index, query_display_name
from shortness
where problem != ''
;
2020-07-03 08:39:43 +00:00
--------------------------------------------------------------------------------
-- various compatibility data formats follow, not related to the main report
-- keep the table in old format so that we can analyze new and old data together
create table queries_old_format engine File(TSVWithNamesAndTypes, 'queries.rep')
2020-09-03 01:42:25 +00:00
as select 0 short, changed_fail, unstable_fail, left, right, diff,
2020-07-03 08:39:43 +00:00
stat_threshold, test, query_display_name query
from queries
;
-- new report for all queries with all metrics (no page yet)
create table all_query_metrics_tsv engine File(TSV, 'report/all-query-metrics.tsv') as
select metric_name, left, right, diff,
floor(left > right ? left / right : right / left, 3),
stat_threshold, test, query_index, query_display_name
from query_metric_stats
2020-07-03 08:39:43 +00:00
left join query_display_names
on query_metric_stats.test = query_display_names.test
and query_metric_stats.query_index = query_display_names.query_index
order by test, query_index;
" 2> >(tee -a report/errors.log 1>&2)
2020-03-24 17:33:18 +00:00
2020-05-25 01:03:21 +00:00
2020-09-25 09:52:09 +00:00
# Prepare source data for metrics and flamegraphs for queries that were profiled
# by perf.py.
2020-03-24 17:33:18 +00:00
for version in {right,left}
2020-05-25 01:03:21 +00:00
do
rm -rf data
2020-05-25 01:42:56 +00:00
clickhouse-local --query "
2020-09-25 09:52:09 +00:00
create view query_profiles as
with 0 as left, 1 as right
2020-09-25 09:52:09 +00:00
select * from file('analyze/query-profiles.tsv', TSV,
'test text, query_index int, query_id text, version UInt8, time float')
where version = $version
;
2020-02-17 19:32:40 +00:00
create view query_display_names as select * from
file('analyze/query-display-names.tsv', TSV,
'test text, query_index int, query_display_name text')
;
2020-02-20 16:28:21 +00:00
2020-03-24 17:33:18 +00:00
create table unstable_query_runs engine File(TSVWithNamesAndTypes,
'unstable-query-runs.$version.rep') as
2020-09-25 09:52:09 +00:00
select query_profiles.test test, query_profiles.query_index query_index,
2020-07-15 13:29:00 +00:00
query_display_name, query_id
2020-09-25 09:52:09 +00:00
from query_profiles
left join query_display_names on
2020-09-25 09:52:09 +00:00
query_profiles.test = query_display_names.test
and query_profiles.query_index = query_display_names.query_index
2020-02-17 19:32:40 +00:00
;
create view query_log as select *
from file('$version-query-log.tsv', TSVWithNamesAndTypes,
'$(cat "$version-query-log.tsv.columns")');
2020-02-20 16:28:21 +00:00
2020-03-24 17:33:18 +00:00
create table unstable_run_metrics engine File(TSVWithNamesAndTypes,
'unstable-run-metrics.$version.rep') as
select
test, query_index, query_id,
ProfileEvents.Values value, ProfileEvents.Names metric
2020-03-24 17:33:18 +00:00
from query_log array join ProfileEvents
join unstable_query_runs using (query_id)
2020-02-17 19:32:40 +00:00
;
2020-03-24 17:33:18 +00:00
create table unstable_run_metrics_2 engine File(TSVWithNamesAndTypes,
'unstable-run-metrics-2.$version.rep') as
select
test, query_index, query_id,
v, n
from (
select
test, query_index, query_id,
2020-03-17 02:41:47 +00:00
['memory_usage', 'read_bytes', 'written_bytes', 'query_duration_ms'] n,
[memory_usage, read_bytes, written_bytes, query_duration_ms] v
2020-03-24 17:33:18 +00:00
from query_log
join unstable_query_runs using (query_id)
)
array join v, n;
create view trace_log as select *
from file('$version-trace-log.tsv', TSVWithNamesAndTypes,
'$(cat "$version-trace-log.tsv.columns")');
2020-05-25 01:03:21 +00:00
create view addresses_src as select addr,
-- Some functions change name between builds, e.g. '__clone' or 'clone' or
-- even '__GI__clone@@GLIBC_2.32'. This breaks differential flame graphs, so
-- filter them out here.
[name, 'clone.S (filtered by script)', 'pthread_cond_timedwait (filtered by script)']
-- this line is a subscript operator of the above array
[1 + multiSearchFirstIndex(name, ['clone.S', 'pthread_cond_timedwait'])] name
from file('$version-addresses.tsv', TSVWithNamesAndTypes,
'$(cat "$version-addresses.tsv.columns")');
create table addresses_join_$version engine Join(any, left, address) as
select addr address, name from addresses_src;
2020-02-20 16:28:21 +00:00
2020-03-24 17:33:18 +00:00
create table unstable_run_traces engine File(TSVWithNamesAndTypes,
'unstable-run-traces.$version.rep') as
select
test, query_index, query_id,
2020-03-24 17:33:18 +00:00
count() value,
2020-05-25 01:03:21 +00:00
joinGet(addresses_join_$version, 'name', arrayJoin(trace))
|| '(' || toString(trace_type) || ')' metric
from trace_log
join unstable_query_runs using query_id
group by test, query_index, query_id, metric
2020-02-17 19:32:40 +00:00
order by count() desc
;
2020-03-24 17:33:18 +00:00
create table metric_devation engine File(TSVWithNamesAndTypes,
2020-05-25 01:03:21 +00:00
'report/metric-deviation.$version.tsv') as
-- first goes the key used to split the file with grep
select test, query_index, query_display_name,
2020-07-31 19:58:18 +00:00
toDecimal64(d, 3) d, q, metric
from (
select
test, query_index,
2020-07-31 19:58:18 +00:00
(q[3] - q[1])/q[2] d,
quantilesExact(0, 0.5, 1)(value) q, metric
from (select * from unstable_run_metrics
union all select * from unstable_run_traces
union all select * from unstable_run_metrics_2) mm
group by test, query_index, metric
2020-07-31 19:58:18 +00:00
having isFinite(d) and d > 0.5 and q[3] > 5
) metrics
2020-05-25 01:42:56 +00:00
left join query_display_names using (test, query_index)
order by test, query_index, d desc
2020-02-17 19:32:40 +00:00
;
2020-02-20 16:28:21 +00:00
2020-05-25 01:03:21 +00:00
create table stacks engine File(TSV, 'report/stacks.$version.tsv') as
2020-02-20 16:28:21 +00:00
select
-- first goes the key used to split the file with grep
2020-05-25 01:03:21 +00:00
test, query_index, trace_type, any(query_display_name),
2020-05-21 23:03:41 +00:00
-- next go the stacks in flamegraph format: 'func1;...;funcN count'
2020-02-20 16:28:21 +00:00
arrayStringConcat(
2020-05-25 01:03:21 +00:00
arrayMap(
addr -> joinGet(addresses_join_$version, 'name', addr),
2020-02-20 16:28:21 +00:00
arrayReverse(trace)
),
';'
) readable_trace,
count() c
2020-03-24 17:33:18 +00:00
from trace_log
2020-02-20 16:28:21 +00:00
join unstable_query_runs using query_id
2020-05-25 01:03:21 +00:00
group by test, query_index, trace_type, trace
order by test, query_index, trace_type, trace
2020-02-20 16:28:21 +00:00
;
2020-07-31 19:58:18 +00:00
" 2> >(tee -a report/errors.log 1>&2) &
2020-03-24 17:33:18 +00:00
done
wait
2020-02-11 15:01:16 +00:00
2020-05-25 01:42:56 +00:00
# Create per-query flamegraphs
2021-04-26 09:39:54 +00:00
touch report/query-files.txt
2020-02-20 16:28:21 +00:00
IFS=$'\n'
2020-03-24 17:33:18 +00:00
for version in {right,left}
2020-02-20 16:28:21 +00:00
do
2020-05-25 01:03:21 +00:00
for query in $(cut -d' ' -f1-4 "report/stacks.$version.tsv" | sort | uniq)
2020-03-24 17:33:18 +00:00
do
query_file=$(echo "$query" | cut -c-120 | sed 's/[/ ]/_/g')
2020-05-21 23:03:41 +00:00
echo "$query_file" >> report/query-files.txt
2020-04-21 18:46:45 +00:00
# Build separate .svg flamegraph for each query.
2020-05-21 23:03:41 +00:00
# -F is somewhat unsafe because it might match not the beginning of the
# string, but this is unlikely and escaping the query for grep is a pain.
2020-05-25 01:03:21 +00:00
grep -F "$query " "report/stacks.$version.tsv" \
| cut -f 5- \
2020-03-24 17:33:18 +00:00
| sed 's/\t/ /g' \
2020-05-25 01:42:56 +00:00
| tee "report/tmp/$query_file.stacks.$version.tsv" \
2020-05-21 23:03:41 +00:00
| ~/fg/flamegraph.pl --hash > "$query_file.$version.svg" &
2020-03-24 17:33:18 +00:00
done
2020-02-20 16:28:21 +00:00
done
wait
unset IFS
2020-05-21 23:03:41 +00:00
# Create differential flamegraphs.
2020-08-19 15:31:13 +00:00
while IFS= read -r query_file
2020-05-21 23:03:41 +00:00
do
2020-05-25 01:42:56 +00:00
~/fg/difffolded.pl "report/tmp/$query_file.stacks.left.tsv" \
"report/tmp/$query_file.stacks.right.tsv" \
| tee "report/tmp/$query_file.stacks.diff.tsv" \
2020-05-21 23:03:41 +00:00
| ~/fg/flamegraph.pl > "$query_file.diff.svg" &
2020-08-19 15:31:13 +00:00
done < report/query-files.txt
2020-05-21 23:03:41 +00:00
wait
2020-05-25 01:42:56 +00:00
# Create per-query files with metrics. Note that the key is different from flamegraphs.
IFS=$'\n'
for version in {right,left}
do
for query in $(cut -d' ' -f1-3 "report/metric-deviation.$version.tsv" | sort | uniq)
do
query_file=$(echo "$query" | cut -c-120 | sed 's/[/ ]/_/g')
# Ditto the above comment about -F.
grep -F "$query " "report/metric-deviation.$version.tsv" \
| cut -f4- > "$query_file.$version.metrics.rep" &
done
done
wait
unset IFS
2020-07-02 07:13:02 +00:00
# Prefer to grep for clickhouse_driver exception messages, but if there are none,
# just show a couple of lines from the log.
for log in *-err.log
do
test=$(basename "$log" "-err.log")
{
2020-09-01 19:05:57 +00:00
# The second grep is a heuristic for error messages like
# "socket.timeout: timed out".
grep -h -m2 -i '\(Exception\|Error\):[^:]' "$log" \
|| grep -h -m2 -i '^[^ ]\+: ' "$log" \
2020-07-02 07:13:02 +00:00
|| head -2 "$log"
} | sed "s/^/$test\t/" >> run-errors.tsv ||:
done
2020-02-07 18:34:24 +00:00
}
2020-02-10 18:37:46 +00:00
2020-06-25 20:19:27 +00:00
function report_metrics
{
2020-09-14 16:16:38 +00:00
build_log_column_definitions
2020-06-25 20:19:27 +00:00
rm -rf metrics ||:
mkdir metrics
2020-07-31 19:58:18 +00:00
clickhouse-local --query "
2020-06-25 20:19:27 +00:00
create view right_async_metric_log as
select * from file('right-async-metric-log.tsv', TSVWithNamesAndTypes,
2020-09-14 16:16:38 +00:00
'$(cat right-async-metric-log.tsv.columns)')
2020-06-25 20:19:27 +00:00
;
-- Use the right log as time reference because it may have higher precision.
create table metrics engine File(TSV, 'metrics/metrics.tsv') as
with (select min(event_time) from right_async_metric_log) as min_time
select name metric, r.event_time - min_time event_time, l.value as left, r.value as right
from right_async_metric_log r
asof join file('left-async-metric-log.tsv', TSVWithNamesAndTypes,
2020-09-14 16:16:38 +00:00
'$(cat left-async-metric-log.tsv.columns)') l
2020-06-25 20:19:27 +00:00
on l.name = r.name and r.event_time <= l.event_time
order by metric, event_time
;
-- Show metrics that have changed
create table changes engine File(TSV, 'metrics/changes.tsv') as
2020-07-31 19:58:18 +00:00
select metric, left, right,
toDecimal64(diff, 3), toDecimal64(times_diff, 3)
from (
select metric, median(left) as left, median(right) as right,
(right - left) / left diff,
if(left > right, left / right, right / left) times_diff
from metrics
group by metric
having abs(diff) > 0.05 and isFinite(diff)
)
2020-06-25 20:19:27 +00:00
order by diff desc
;
2020-08-19 15:31:13 +00:00
" 2> >(tee -a metrics/errors.log 1>&2)
2020-06-25 20:19:27 +00:00
IFS=$'\n'
for prefix in $(cut -f1 "metrics/metrics.tsv" | sort | uniq)
do
file="metrics/$prefix.tsv"
grep "^$prefix " "metrics/metrics.tsv" | cut -f2- > "$file"
gnuplot -e "
set datafile separator '\t';
set terminal png size 960,540;
set xtics time format '%tH:%tM';
set title '$prefix' noenhanced offset 0,-3;
set key left top;
plot
'$file' using 1:2 with lines title 'Left'
, '$file' using 1:3 with lines title 'Right'
;
" \
| convert - -filter point -resize "200%" "metrics/$prefix.png" &
done
wait
unset IFS
}
function upload_results
{
if ! [ -v CHPC_DATABASE_URL ]
then
echo Database for test results is not specified, will not upload them.
return 0
fi
2020-11-04 01:19:53 +00:00
set +x # Don't show password in the log
2021-04-22 16:46:54 +00:00
client=(clickhouse-client
# Surprisingly, clickhouse-client doesn't understand --host 127.0.0.1:9000
# so I have to extract host and port with clickhouse-local. I tried to use
# Poco URI parser to support this in the client, but it's broken and can't
# parse host:port.
$(clickhouse-local --query "with '${CHPC_DATABASE_URL}' as url select '--host ' || domain(url) || ' --port ' || toString(port(url)) format TSV")
--secure
--user "${CHPC_DATABASE_USER}"
--password "${CHPC_DATABASE_PASSWORD}"
--config "right/config/client_config.xml"
--database perftest
--date_time_input_format=best_effort)
"${client[@]}" --query "
2020-11-06 10:45:42 +00:00
insert into query_metrics_v2
2020-11-04 01:19:53 +00:00
select
toDate(event_time) event_date,
toDateTime('$(cd right/ch && git show -s --format=%ci "$SHA_TO_TEST" | cut -d' ' -f-2)') event_time,
$PR_TO_TEST pr_number,
'$REF_SHA' old_sha,
'$SHA_TO_TEST' new_sha,
test,
query_index,
query_display_name,
metric_name,
old_value,
new_value,
diff,
stat_threshold
from input('metric_name text, old_value float, new_value float, diff float,
ratio_display_text text, stat_threshold float,
test text, query_index int, query_display_name text')
settings date_time_input_format='best_effort'
format TSV
settings date_time_input_format='best_effort'
" < report/all-query-metrics.tsv # Don't leave whitespace after INSERT: https://github.com/ClickHouse/ClickHouse/issues/16652
2021-04-22 16:46:54 +00:00
# Upload some run attributes. I use this weird form because it is the same
# form that can be used for historical data when you only have compare.log.
cat compare.log \
| sed -n '
2021-04-22 17:18:21 +00:00
s/.*Model name:[[:space:]]\+\(.*\)$/metric lscpu-model-name \1/p;
s/.*L1d cache:[[:space:]]\+\(.*\)$/metric lscpu-l1d-cache \1/p;
s/.*L1i cache:[[:space:]]\+\(.*\)$/metric lscpu-l1i-cache \1/p;
s/.*L2 cache:[[:space:]]\+\(.*\)$/metric lscpu-l2-cache \1/p;
s/.*L3 cache:[[:space:]]\+\(.*\)$/metric lscpu-l3-cache \1/p;
2021-04-22 16:46:54 +00:00
s/.*left_sha=\(.*\)$/old-sha \1/p;
s/.*right_sha=\(.*\)/new-sha \1/p' \
| awk '
BEGIN { FS = "\t"; OFS = "\t" }
/^old-sha/ { old_sha=$2 }
/^new-sha/ { new_sha=$2 }
/^metric/ { print old_sha, new_sha, $2, $3 }' \
| "${client[@]}" --query "INSERT INTO run_attributes_v1 FORMAT TSV"
# Grepping numactl results from log is too crazy, I'll just call it again.
"${client[@]}" --query "INSERT INTO run_attributes_v1 FORMAT TSV" <<EOF
2021-04-26 09:39:54 +00:00
$REF_SHA $SHA_TO_TEST $(numactl --show | sed -n 's/^cpubind:[[:space:]]\+/numactl-cpubind /p')
$REF_SHA $SHA_TO_TEST $(numactl --hardware | sed -n 's/^available:[[:space:]]\+/numactl-available /p')
EOF
set -x
}
2020-04-17 15:47:01 +00:00
# Check that local and client are in PATH
clickhouse-local --version > /dev/null
clickhouse-client --version > /dev/null
2020-02-10 18:37:46 +00:00
case "$stage" in
"")
;&
2020-02-14 19:11:46 +00:00
"configure")
time configure
;&
"restart")
2020-10-22 12:52:43 +00:00
numactl --show ||:
numactl --hardware ||:
lscpu ||:
2020-10-21 15:06:42 +00:00
dmidecode -t 4 ||:
2020-02-14 19:11:46 +00:00
time restart
;&
"run_tests")
2020-03-17 16:37:09 +00:00
# Ignore the errors to collect the log and build at least some report, anyway
2020-02-25 19:51:09 +00:00
time run_tests ||:
2020-02-14 19:11:46 +00:00
;&
"get_profiles")
2020-06-25 20:19:27 +00:00
# Check for huge pages.
cat /sys/kernel/mm/transparent_hugepage/enabled > thp-enabled.txt ||:
cat /proc/meminfo > meminfo.txt ||:
for pid in $(pgrep -f clickhouse-server)
do
cat "/proc/$pid/smaps" > "$pid-smaps.txt" ||:
done
# We had a bug where getting profiles froze sometimes, so try to save some
# logs if this happens again. Give the servers some time to collect all info,
# then trace and kill. Start in a subshell, so that both function don't
# interfere with each other's jobs through `wait`. Also make the subshell
# have its own process group, so that we can then kill it with all its child
# processes. Somehow it doesn't kill the children by itself when dying.
2020-04-21 21:53:13 +00:00
set -m
( get_profiles_watchdog ) &
watchdog_pid=$!
set +m
# Check that the watchdog started OK.
kill -0 $watchdog_pid
2020-02-28 16:22:07 +00:00
# If the tests fail with OOM or something, still try to restart the servers
# to collect the logs. Prefer not to restart, because addresses might change
2020-04-21 21:53:13 +00:00
# and we won't be able to process trace_log data. Start in a subshell, so that
# it doesn't interfere with the watchdog through `wait`.
2020-09-23 08:21:55 +00:00
( get_profiles || { restart && get_profiles ; } ) ||:
2020-04-21 21:53:13 +00:00
# Kill the whole process group, because somehow when the subshell is killed,
# the sleep inside remains alive and orphaned.
while env kill -- -$watchdog_pid ; do sleep 1; done
2020-03-03 10:47:32 +00:00
# Stop the servers to free memory for the subsequent query analysis.
while killall clickhouse; do echo . ; sleep 1 ; done
echo Servers stopped.
2020-02-10 18:37:46 +00:00
;&
2020-02-27 19:43:43 +00:00
"analyze_queries")
2020-04-28 14:26:49 +00:00
time analyze_queries ||:
2020-02-27 19:43:43 +00:00
;&
2020-02-10 18:37:46 +00:00
"report")
2020-04-30 08:36:33 +00:00
time report ||:
2020-06-25 20:19:27 +00:00
;&
"report_metrics")
time report_metrics ||:
2020-07-31 19:58:18 +00:00
cat metrics/errors.log >> report/errors.log ||:
2020-06-25 20:19:27 +00:00
;&
"report_html")
2020-04-28 07:45:35 +00:00
time "$script_dir/report.py" --report=all-queries > all-queries.html 2> >(tee -a report/errors.log 1>&2) ||:
2020-03-17 16:37:09 +00:00
time "$script_dir/report.py" > report.html
2020-02-10 18:37:46 +00:00
;&
"upload_results")
time upload_results ||:
;&
2020-02-10 18:37:46 +00:00
esac
2020-04-21 21:53:13 +00:00
# Print some final debug info to help debug Weirdness, of which there is plenty.
jobs
pstree -apgT
2021-04-26 09:41:29 +00:00