#!/bin/bash # fail on errors, verbose and export all env variables set -e -x -a # shellcheck disable=SC1091 source /setup_export_logs.sh # shellcheck source=../stateless/stress_tests.lib source /repo/tests/docker_scripts/stress_tests.lib # Avoid overlaps with previous runs dmesg --clear # fail on errors, verbose and export all env variables set -e -x -a USE_DATABASE_REPLICATED=${USE_DATABASE_REPLICATED:=0} USE_SHARED_CATALOG=${USE_SHARED_CATALOG:=0} # Choose random timezone for this test run. # # NOTE: that clickhouse-test will randomize session_timezone by itself as well # (it will choose between default server timezone and something specific). TZ="$(rg -v '#' /usr/share/zoneinfo/zone.tab | awk '{print $3}' | shuf | head -n1)" echo "Chosen random timezone $TZ" ln -snf "/usr/share/zoneinfo/$TZ" /etc/localtime && echo "$TZ" > /etc/timezone dpkg -i package_folder/clickhouse-common-static_*.deb dpkg -i package_folder/clickhouse-common-static-dbg_*.deb dpkg -i package_folder/clickhouse-odbc-bridge_*.deb dpkg -i package_folder/clickhouse-library-bridge_*.deb dpkg -i package_folder/clickhouse-server_*.deb dpkg -i package_folder/clickhouse-client_*.deb echo "$BUGFIX_VALIDATE_CHECK" # Check that the tools are available under short names if [[ -z "$BUGFIX_VALIDATE_CHECK" ]]; then ch --query "SELECT 1" || exit 1 chl --query "SELECT 1" || exit 1 chc --version || exit 1 fi ln -sf /repo/tests/clickhouse-test /usr/bin/clickhouse-test export CLICKHOUSE_GRPC_CLIENT="/repo/utils/grpc-client/clickhouse-grpc-client.py" # shellcheck disable=SC1091 source /repo/tests/docker_scripts/attach_gdb.lib # shellcheck disable=SC1091 source /repo/tests/docker_scripts/utils.lib # install test configs /repo/tests/config/install.sh /repo/tests/docker_scripts/setup_minio.sh stateless /repo/tests/docker_scripts/setup_hdfs_minicluster.sh config_logs_export_cluster /etc/clickhouse-server/config.d/system_logs_export.yaml if [[ -n "$BUGFIX_VALIDATE_CHECK" ]] && [[ "$BUGFIX_VALIDATE_CHECK" -eq 1 ]]; then sudo sed -i "/1<\/use_compression>/d" /etc/clickhouse-server/config.d/zookeeper.xml # it contains some new settings, but we can safely remove it rm /etc/clickhouse-server/config.d/handlers.yaml rm /etc/clickhouse-server/users.d/s3_cache_new.xml rm /etc/clickhouse-server/config.d/zero_copy_destructive_operations.xml function remove_keeper_config() { sudo sed -i "/<$1>$2<\/$1>/d" /etc/clickhouse-server/config.d/keeper_port.xml } # commit_logs_cache_size_threshold setting doesn't exist on some older versions remove_keeper_config "commit_logs_cache_size_threshold" "[[:digit:]]\+" remove_keeper_config "latest_logs_cache_size_threshold" "[[:digit:]]\+" fi export IS_FLAKY_CHECK=0 # Export NUM_TRIES so python scripts will see its value as env variable export NUM_TRIES # For flaky check we also enable thread fuzzer if [ "$NUM_TRIES" -gt "1" ]; then export IS_FLAKY_CHECK=1 export THREAD_FUZZER_CPU_TIME_PERIOD_US=1000 export THREAD_FUZZER_SLEEP_PROBABILITY=0.1 export THREAD_FUZZER_SLEEP_TIME_US_MAX=100000 export THREAD_FUZZER_pthread_mutex_lock_BEFORE_MIGRATE_PROBABILITY=1 export THREAD_FUZZER_pthread_mutex_lock_AFTER_MIGRATE_PROBABILITY=1 export THREAD_FUZZER_pthread_mutex_unlock_BEFORE_MIGRATE_PROBABILITY=1 export THREAD_FUZZER_pthread_mutex_unlock_AFTER_MIGRATE_PROBABILITY=1 export THREAD_FUZZER_pthread_mutex_lock_BEFORE_SLEEP_PROBABILITY=0.001 export THREAD_FUZZER_pthread_mutex_lock_AFTER_SLEEP_PROBABILITY=0.001 export THREAD_FUZZER_pthread_mutex_unlock_BEFORE_SLEEP_PROBABILITY=0.001 export THREAD_FUZZER_pthread_mutex_unlock_AFTER_SLEEP_PROBABILITY=0.001 export THREAD_FUZZER_pthread_mutex_lock_BEFORE_SLEEP_TIME_US_MAX=10000 export THREAD_FUZZER_pthread_mutex_lock_AFTER_SLEEP_TIME_US_MAX=10000 export THREAD_FUZZER_pthread_mutex_unlock_BEFORE_SLEEP_TIME_US_MAX=10000 export THREAD_FUZZER_pthread_mutex_unlock_AFTER_SLEEP_TIME_US_MAX=10000 mkdir -p /var/run/clickhouse-server fi # simplest way to forward env variables to server sudo -E -u clickhouse /usr/bin/clickhouse-server --config /etc/clickhouse-server/config.xml --daemon --pid-file /var/run/clickhouse-server/clickhouse-server.pid if [[ "$USE_DATABASE_REPLICATED" -eq 1 ]]; then sudo sed -i "s|/var/lib/clickhouse/filesystem_caches/|/var/lib/clickhouse/filesystem_caches_1/|" /etc/clickhouse-server1/config.d/filesystem_caches_path.xml sudo sed -i "s|/var/lib/clickhouse/filesystem_caches/|/var/lib/clickhouse/filesystem_caches_2/|" /etc/clickhouse-server2/config.d/filesystem_caches_path.xml sudo sed -i "s|/var/lib/clickhouse/filesystem_caches/|/var/lib/clickhouse/filesystem_caches_1/|" /etc/clickhouse-server1/config.d/filesystem_caches_path.xml sudo sed -i "s|/var/lib/clickhouse/filesystem_caches/|/var/lib/clickhouse/filesystem_caches_2/|" /etc/clickhouse-server2/config.d/filesystem_caches_path.xml mkdir -p /var/run/clickhouse-server1 sudo chown clickhouse:clickhouse /var/run/clickhouse-server1 sudo -E -u clickhouse /usr/bin/clickhouse server --config /etc/clickhouse-server1/config.xml --daemon \ --pid-file /var/run/clickhouse-server1/clickhouse-server.pid \ -- --path /var/lib/clickhouse1/ --logger.stderr /var/log/clickhouse-server/stderr1.log \ --logger.log /var/log/clickhouse-server/clickhouse-server1.log --logger.errorlog /var/log/clickhouse-server/clickhouse-server1.err.log \ --tcp_port 19000 --tcp_port_secure 19440 --http_port 18123 --https_port 18443 --interserver_http_port 19009 --tcp_with_proxy_port 19010 \ --mysql_port 19004 --postgresql_port 19005 \ --keeper_server.tcp_port 19181 --keeper_server.server_id 2 \ --prometheus.port 19988 \ --macros.replica r2 # It doesn't work :( mkdir -p /var/run/clickhouse-server2 sudo chown clickhouse:clickhouse /var/run/clickhouse-server2 sudo -E -u clickhouse /usr/bin/clickhouse server --config /etc/clickhouse-server2/config.xml --daemon \ --pid-file /var/run/clickhouse-server2/clickhouse-server.pid \ -- --path /var/lib/clickhouse2/ --logger.stderr /var/log/clickhouse-server/stderr2.log \ --logger.log /var/log/clickhouse-server/clickhouse-server2.log --logger.errorlog /var/log/clickhouse-server/clickhouse-server2.err.log \ --tcp_port 29000 --tcp_port_secure 29440 --http_port 28123 --https_port 28443 --interserver_http_port 29009 --tcp_with_proxy_port 29010 \ --mysql_port 29004 --postgresql_port 29005 \ --keeper_server.tcp_port 29181 --keeper_server.server_id 3 \ --prometheus.port 29988 \ --macros.shard s2 # It doesn't work :( fi if [[ "$USE_SHARED_CATALOG" -eq 1 ]]; then sudo cat /etc/clickhouse-server1/config.d/filesystem_caches_path.xml \ | sed "s|/var/lib/clickhouse/filesystem_caches/|/var/lib/clickhouse/filesystem_caches_1/|" \ > /etc/clickhouse-server1/config.d/filesystem_caches_path.xml.tmp mv /etc/clickhouse-server1/config.d/filesystem_caches_path.xml.tmp /etc/clickhouse-server1/config.d/filesystem_caches_path.xml sudo cat /etc/clickhouse-server1/config.d/filesystem_caches_path.xml \ | sed "s|/var/lib/clickhouse/filesystem_caches/|/var/lib/clickhouse/filesystem_caches_1/|" \ > /etc/clickhouse-server1/config.d/filesystem_caches_path.xml.tmp mv /etc/clickhouse-server1/config.d/filesystem_caches_path.xml.tmp /etc/clickhouse-server1/config.d/filesystem_caches_path.xml mkdir -p /var/run/clickhouse-server1 sudo chown clickhouse:clickhouse /var/run/clickhouse-server1 sudo -E -u clickhouse /usr/bin/clickhouse server --config /etc/clickhouse-server1/config.xml --daemon \ --pid-file /var/run/clickhouse-server1/clickhouse-server.pid \ -- --path /var/lib/clickhouse1/ --logger.stderr /var/log/clickhouse-server/stderr1.log \ --logger.log /var/log/clickhouse-server/clickhouse-server1.log --logger.errorlog /var/log/clickhouse-server/clickhouse-server1.err.log \ --tcp_port 19000 --tcp_port_secure 19440 --http_port 18123 --https_port 18443 --interserver_http_port 19009 --tcp_with_proxy_port 19010 \ --mysql_port 19004 --postgresql_port 19005 \ --keeper_server.tcp_port 19181 --keeper_server.server_id 2 \ --prometheus.port 19988 \ --macros.replica r2 # It doesn't work :( fi # Wait for the server to start, but not for too long. for _ in {1..100} do clickhouse-client --query "SELECT 1" && break sleep 1 done setup_logs_replication attach_gdb_to_clickhouse # create tables for minio log webhooks clickhouse-client --query "CREATE TABLE minio_audit_logs ( log String, event_time DateTime64(9) MATERIALIZED parseDateTime64BestEffortOrZero(trim(BOTH '\"' FROM JSONExtractRaw(log, 'time')), 9, 'UTC') ) ENGINE = MergeTree ORDER BY tuple()" clickhouse-client --query "CREATE TABLE minio_server_logs ( log String, event_time DateTime64(9) MATERIALIZED parseDateTime64BestEffortOrZero(trim(BOTH '\"' FROM JSONExtractRaw(log, 'time')), 9, 'UTC') ) ENGINE = MergeTree ORDER BY tuple()" # create minio log webhooks for both audit and server logs # use async inserts to avoid creating too many parts ./mc admin config set clickminio logger_webhook:ch_server_webhook endpoint="http://localhost:8123/?async_insert=1&wait_for_async_insert=0&async_insert_busy_timeout_min_ms=5000&async_insert_busy_timeout_max_ms=5000&async_insert_max_query_number=1000&async_insert_max_data_size=10485760&query=INSERT%20INTO%20minio_server_logs%20FORMAT%20LineAsString" queue_size=1000000 batch_size=500 ./mc admin config set clickminio audit_webhook:ch_audit_webhook endpoint="http://localhost:8123/?async_insert=1&wait_for_async_insert=0&async_insert_busy_timeout_min_ms=5000&async_insert_busy_timeout_max_ms=5000&async_insert_max_query_number=1000&async_insert_max_data_size=10485760&query=INSERT%20INTO%20minio_audit_logs%20FORMAT%20LineAsString" queue_size=1000000 batch_size=500 max_retries=100 retry=1 while [ $retry -le $max_retries ]; do echo "clickminio restart attempt $retry:" output=$(./mc admin service restart clickminio --wait --json 2>&1 | jq -r .status) echo "Output of restart status: $output" expected_output="success success" if [ "$output" = "$expected_output" ]; then echo "Restarted clickminio successfully." break fi sleep 1 retry=$((retry + 1)) done if [ $retry -gt $max_retries ]; then echo "Failed to restart clickminio after $max_retries attempts." fi ./mc admin trace clickminio > /test_output/minio.log & MC_ADMIN_PID=$! function fn_exists() { declare -F "$1" > /dev/null; } # FIXME: to not break old builds, clean on 2023-09-01 function try_run_with_retry() { local total_retries="$1" shift if fn_exists run_with_retry; then run_with_retry "$total_retries" "$@" else "$@" fi } function run_tests() { set -x # We can have several additional options so we pass them as array because it is more ideologically correct. read -ra ADDITIONAL_OPTIONS <<< "${ADDITIONAL_OPTIONS:-}" HIGH_LEVEL_COVERAGE=YES # Use random order in flaky check if [ "$NUM_TRIES" -gt "1" ]; then ADDITIONAL_OPTIONS+=('--order=random') HIGH_LEVEL_COVERAGE=NO fi if [[ -n "$USE_S3_STORAGE_FOR_MERGE_TREE" ]] && [[ "$USE_S3_STORAGE_FOR_MERGE_TREE" -eq 1 ]]; then ADDITIONAL_OPTIONS+=('--s3-storage') fi if [[ -n "$USE_AZURE_STORAGE_FOR_MERGE_TREE" ]] && [[ "$USE_AZURE_STORAGE_FOR_MERGE_TREE" -eq 1 ]]; then # to disable the same tests ADDITIONAL_OPTIONS+=('--azure-blob-storage') # azurite is slow, but with these two settings it can be super slow ADDITIONAL_OPTIONS+=('--no-random-settings') ADDITIONAL_OPTIONS+=('--no-random-merge-tree-settings') fi if [[ "$USE_SHARED_CATALOG" -eq 1 ]]; then ADDITIONAL_OPTIONS+=('--shared-catalog') fi if [[ "$USE_DISTRIBUTED_CACHE" -eq 1 ]]; then ADDITIONAL_OPTIONS+=('--distributed-cache') fi if [[ "$USE_DATABASE_REPLICATED" -eq 1 ]]; then ADDITIONAL_OPTIONS+=('--replicated-database') # Too many tests fail for DatabaseReplicated in parallel. ADDITIONAL_OPTIONS+=('--jobs') ADDITIONAL_OPTIONS+=('3') elif [[ 1 == $(clickhouse-client --query "SELECT value LIKE '%SANITIZE_COVERAGE%' FROM system.build_options WHERE name = 'CXX_FLAGS'") ]]; then # Coverage on a per-test basis could only be collected sequentially. # Do not set the --jobs parameter. echo "Running tests with coverage collection." else # All other configurations are OK. ADDITIONAL_OPTIONS+=('--jobs') ADDITIONAL_OPTIONS+=('8') fi if [[ -n "$RUN_BY_HASH_NUM" ]] && [[ -n "$RUN_BY_HASH_TOTAL" ]]; then ADDITIONAL_OPTIONS+=('--run-by-hash-num') ADDITIONAL_OPTIONS+=("$RUN_BY_HASH_NUM") ADDITIONAL_OPTIONS+=('--run-by-hash-total') ADDITIONAL_OPTIONS+=("$RUN_BY_HASH_TOTAL") HIGH_LEVEL_COVERAGE=NO fi if [[ -n "$USE_DATABASE_ORDINARY" ]] && [[ "$USE_DATABASE_ORDINARY" -eq 1 ]]; then ADDITIONAL_OPTIONS+=('--db-engine=Ordinary') fi if [[ "${HIGH_LEVEL_COVERAGE}" = "YES" ]]; then ADDITIONAL_OPTIONS+=('--report-coverage') fi ADDITIONAL_OPTIONS+=('--report-logs-stats') try_run_with_retry 10 clickhouse-client -q "insert into system.zookeeper (name, path, value) values ('auxiliary_zookeeper2', '/test/chroot/', '')" set +e TEST_ARGS=( --testname --shard --zookeeper --check-zookeeper-session --hung-check --print-time --no-drop-if-fail --capture-client-stacktrace --queries "/repo/tests/queries" --test-runs "$NUM_TRIES" "${ADDITIONAL_OPTIONS[@]}" ) clickhouse-test "${TEST_ARGS[@]}" 2>&1 \ | ts '%Y-%m-%d %H:%M:%S' \ | tee -a test_output/test_result.txt set -e } export -f run_tests if [ "$NUM_TRIES" -gt "1" ]; then # We don't run tests with Ordinary database in PRs, only in master. # So run new/changed tests with Ordinary at least once in flaky check. NUM_TRIES=1 USE_DATABASE_ORDINARY=1 run_tests \ | sed 's/All tests have finished/Redacted: a message about tests finish is deleted/' | sed 's/No tests were run/Redacted: a message about no tests run is deleted/' ||: fi run_tests ||: echo "Files in current directory" ls -la ./ echo "Files in root directory" ls -la / /repo/tests/docker_scripts/process_functional_tests_result.py || echo -e "failure\tCannot parse results" > /test_output/check_status.tsv clickhouse-client -q "system flush logs" ||: # stop logs replication to make it possible to dump logs tables via clickhouse-local stop_logs_replication # Try to get logs while server is running failed_to_save_logs=0 for table in query_log zookeeper_log trace_log transactions_info_log metric_log blob_storage_log error_log do if ! clickhouse-client -q "select * from system.$table into outfile '/test_output/$table.tsv.zst' format TSVWithNamesAndTypes"; then failed_to_save_logs=1 fi if [[ "$USE_DATABASE_REPLICATED" -eq 1 ]]; then if ! clickhouse-client --port 19000 -q "select * from system.$table into outfile '/test_output/$table.1.tsv.zst' format TSVWithNamesAndTypes"; then failed_to_save_logs=1 fi if ! clickhouse-client --port 29000 -q "select * from system.$table into outfile '/test_output/$table.2.tsv.zst' format TSVWithNamesAndTypes"; then failed_to_save_logs=1 fi fi if [[ "$USE_SHARED_CATALOG" -eq 1 ]]; then if ! clickhouse-client --port 29000 -q "select * from system.$table into outfile '/test_output/$table.2.tsv.zst' format TSVWithNamesAndTypes"; then failed_to_save_logs=1 fi fi done # collect minio audit and server logs # wait for minio to flush its batch if it has any sleep 1 clickhouse-client -q "SYSTEM FLUSH ASYNC INSERT QUEUE" clickhouse-client --max_block_size 8192 --max_memory_usage 10G --max_threads 1 --max_result_bytes 0 --max_result_rows 0 --max_rows_to_read 0 --max_bytes_to_read 0 -q "SELECT log FROM minio_audit_logs ORDER BY event_time INTO OUTFILE '/test_output/minio_audit_logs.jsonl.zst' FORMAT JSONEachRow" clickhouse-client --max_block_size 8192 --max_memory_usage 10G --max_threads 1 --max_result_bytes 0 --max_result_rows 0 --max_rows_to_read 0 --max_bytes_to_read 0 -q "SELECT log FROM minio_server_logs ORDER BY event_time INTO OUTFILE '/test_output/minio_server_logs.jsonl.zst' FORMAT JSONEachRow" # Stop server so we can safely read data with clickhouse-local. # Why do we read data with clickhouse-local? # Because it's the simplest way to read it when server has crashed. sudo clickhouse stop ||: if [[ "$USE_DATABASE_REPLICATED" -eq 1 ]]; then sudo clickhouse stop --pid-path /var/run/clickhouse-server1 ||: sudo clickhouse stop --pid-path /var/run/clickhouse-server2 ||: fi if [[ "$USE_SHARED_CATALOG" -eq 1 ]]; then sudo clickhouse stop --pid-path /var/run/clickhouse-server1 ||: fi # Kill minio admin client to stop collecting logs kill $MC_ADMIN_PID rg -Fa "" /var/log/clickhouse-server/clickhouse-server.log ||: rg -A50 -Fa "============" /var/log/clickhouse-server/stderr.log ||: zstd --threads=0 < /var/log/clickhouse-server/clickhouse-server.log > /test_output/clickhouse-server.log.zst & data_path_config="--path=/var/lib/clickhouse/" if [[ -n "$USE_S3_STORAGE_FOR_MERGE_TREE" ]] && [[ "$USE_S3_STORAGE_FOR_MERGE_TREE" -eq 1 ]]; then # We need s3 storage configuration (but it's more likely that clickhouse-local will fail for some reason) data_path_config="--config-file=/etc/clickhouse-server/config.xml" fi # If server crashed dump system logs with clickhouse-local if [ $failed_to_save_logs -ne 0 ]; then # Compress tables. # # NOTE: # - that due to tests with s3 storage we cannot use /var/lib/clickhouse/data # directly # - even though ci auto-compress some files (but not *.tsv) it does this only # for files >64MB, we want this files to be compressed explicitly for table in query_log zookeeper_log trace_log transactions_info_log metric_log blob_storage_log error_log do clickhouse-local "$data_path_config" --only-system-tables --stacktrace -q "select * from system.$table format TSVWithNamesAndTypes" | zstd --threads=0 > /test_output/$table.tsv.zst ||: if [[ "$USE_DATABASE_REPLICATED" -eq 1 ]]; then clickhouse-local --path /var/lib/clickhouse1/ --only-system-tables --stacktrace -q "select * from system.$table format TSVWithNamesAndTypes" | zstd --threads=0 > /test_output/$table.1.tsv.zst ||: clickhouse-local --path /var/lib/clickhouse2/ --only-system-tables --stacktrace -q "select * from system.$table format TSVWithNamesAndTypes" | zstd --threads=0 > /test_output/$table.2.tsv.zst ||: fi if [[ "$USE_SHARED_CATALOG" -eq 1 ]]; then clickhouse-local --path /var/lib/clickhouse1/ --only-system-tables --stacktrace -q "select * from system.$table format TSVWithNamesAndTypes" | zstd --threads=0 > /test_output/$table.1.tsv.zst ||: fi done fi # Also export trace log in flamegraph-friendly format. for trace_type in CPU Memory Real do clickhouse-local "$data_path_config" --only-system-tables -q " select arrayStringConcat((arrayMap(x -> concat(splitByChar('/', addressToLine(x))[-1], '#', demangle(addressToSymbol(x)) ), trace)), ';') AS stack, count(*) AS samples from system.trace_log where trace_type = '$trace_type' group by trace order by samples desc settings allow_introspection_functions = 1 format TabSeparated" \ | zstd --threads=0 > "/test_output/trace-log-$trace_type-flamegraph.tsv.zst" ||: done # Grep logs for sanitizer asserts, crashes and other critical errors check_logs_for_critical_errors # Compressed (FIXME: remove once only github actions will be left) rm /var/log/clickhouse-server/clickhouse-server.log mv /var/log/clickhouse-server/stderr.log /test_output/ ||: if [[ -n "$WITH_COVERAGE" ]] && [[ "$WITH_COVERAGE" -eq 1 ]]; then tar --zstd -chf /test_output/clickhouse_coverage.tar.zst /profraw ||: fi tar -chf /test_output/coordination.tar /var/lib/clickhouse/coordination ||: rm -rf /var/lib/clickhouse/data/system/*/ tar -chf /test_output/store.tar /var/lib/clickhouse/store ||: tar -chf /test_output/metadata.tar /var/lib/clickhouse/metadata/*.sql ||: if [[ "$USE_DATABASE_REPLICATED" -eq 1 ]]; then rg -Fa "" /var/log/clickhouse-server/clickhouse-server1.log ||: rg -Fa "" /var/log/clickhouse-server/clickhouse-server2.log ||: zstd --threads=0 < /var/log/clickhouse-server/clickhouse-server1.log > /test_output/clickhouse-server1.log.zst ||: zstd --threads=0 < /var/log/clickhouse-server/clickhouse-server2.log > /test_output/clickhouse-server2.log.zst ||: mv /var/log/clickhouse-server/stderr1.log /test_output/ ||: mv /var/log/clickhouse-server/stderr2.log /test_output/ ||: tar -chf /test_output/coordination1.tar /var/lib/clickhouse1/coordination ||: tar -chf /test_output/coordination2.tar /var/lib/clickhouse2/coordination ||: fi if [[ "$USE_SHARED_CATALOG" -eq 1 ]]; then rg -Fa "" /var/log/clickhouse-server/clickhouse-server1.log ||: zstd --threads=0 < /var/log/clickhouse-server/clickhouse-server1.log > /test_output/clickhouse-server1.log.zst ||: mv /var/log/clickhouse-server/stderr1.log /test_output/ ||: tar -chf /test_output/coordination1.tar /var/lib/clickhouse1/coordination ||: fi collect_core_dumps