Merge pull request ClickHouse#79633 from azat/tests/fix-filesystem-cache-tests

tests: fix flakiness of tests for system.filesystem_cache

Author: alexey-milovidov

tests/.gitignore

@@ -6,3 +6,5 @@ test_data
 
 /queries/0_stateless/*.gen.sql
 /queries/0_stateless/*.gen.reference
+
+!*.lib

tests/docker_scripts/stateless_runner.sh

@@ -243,6 +243,10 @@ function prepare_stateful_data() {
             ENGINE = CollapsingMergeTree(Sign) PARTITION BY toYYYYMM(StartDate) ORDER BY (CounterID, StartDate, intHash32(UserID), VisitID)
             SAMPLE BY intHash32(UserID) SETTINGS index_granularity = 8192, storage_policy='s3_cache'"
 
+        # Avoid locking filesystem cache
+        clickhouse-client --query "SYSTEM STOP MERGES test.hits"
+        clickhouse-client --query "SYSTEM STOP MERGES test.visits"
+
         clickhouse-client --max_execution_time 600 --max_memory_usage 25G --query "INSERT INTO test.hits SELECT * FROM datasets.hits_v1 SETTINGS enable_filesystem_cache_on_write_operations=0, max_insert_threads=16"
         clickhouse-client --max_execution_time 600 --max_memory_usage 25G --query "INSERT INTO test.visits SELECT * FROM datasets.visits_v1 SETTINGS enable_filesystem_cache_on_write_operations=0, max_insert_threads=16"
         clickhouse-client --query "DROP TABLE datasets.visits_v1 SYNC"
@@ -252,6 +256,8 @@ function prepare_stateful_data() {
         clickhouse-client --query "RENAME TABLE datasets.visits_v1 TO test.visits"
     fi
     clickhouse-client --query "CREATE TABLE test.hits_s3  (WatchID UInt64, JavaEnable UInt8, Title String, GoodEvent Int16, EventTime DateTime, EventDate Date, CounterID UInt32, ClientIP UInt32, ClientIP6 FixedString(16), RegionID UInt32, UserID UInt64, CounterClass Int8, OS UInt8, UserAgent UInt8, URL String, Referer String, URLDomain String, RefererDomain String, Refresh UInt8, IsRobot UInt8, RefererCategories Array(UInt16), URLCategories Array(UInt16), URLRegions Array(UInt32), RefererRegions Array(UInt32), ResolutionWidth UInt16, ResolutionHeight UInt16, ResolutionDepth UInt8, FlashMajor UInt8, FlashMinor UInt8, FlashMinor2 String, NetMajor UInt8, NetMinor UInt8, UserAgentMajor UInt16, UserAgentMinor FixedString(2), CookieEnable UInt8, JavascriptEnable UInt8, IsMobile UInt8, MobilePhone UInt8, MobilePhoneModel String, Params String, IPNetworkID UInt32, TraficSourceID Int8, SearchEngineID UInt16, SearchPhrase String, AdvEngineID UInt8, IsArtifical UInt8, WindowClientWidth UInt16, WindowClientHeight UInt16, ClientTimeZone Int16, ClientEventTime DateTime, SilverlightVersion1 UInt8, SilverlightVersion2 UInt8, SilverlightVersion3 UInt32, SilverlightVersion4 UInt16, PageCharset String, CodeVersion UInt32, IsLink UInt8, IsDownload UInt8, IsNotBounce UInt8, FUniqID UInt64, HID UInt32, IsOldCounter UInt8, IsEvent UInt8, IsParameter UInt8, DontCountHits UInt8, WithHash UInt8, HitColor FixedString(1), UTCEventTime DateTime, Age UInt8, Sex UInt8, Income UInt8, Interests UInt16, Robotness UInt8, GeneralInterests Array(UInt16), RemoteIP UInt32, RemoteIP6 FixedString(16), WindowName Int32, OpenerName Int32, HistoryLength Int16, BrowserLanguage FixedString(2), BrowserCountry FixedString(2), SocialNetwork String, SocialAction String, HTTPError UInt16, SendTiming Int32, DNSTiming Int32, ConnectTiming Int32, ResponseStartTiming Int32, ResponseEndTiming Int32, FetchTiming Int32, RedirectTiming Int32, DOMInteractiveTiming Int32, DOMContentLoadedTiming Int32, DOMCompleteTiming Int32, LoadEventStartTiming Int32, LoadEventEndTiming Int32, NSToDOMContentLoadedTiming Int32, FirstPaintTiming Int32, RedirectCount Int8, SocialSourceNetworkID UInt8, SocialSourcePage String, ParamPrice Int64, ParamOrderID String, ParamCurrency FixedString(3), ParamCurrencyID UInt16, GoalsReached Array(UInt32), OpenstatServiceName String, OpenstatCampaignID String, OpenstatAdID String, OpenstatSourceID String, UTMSource String, UTMMedium String, UTMCampaign String, UTMContent String, UTMTerm String, FromTag String, HasGCLID UInt8, RefererHash UInt64, URLHash UInt64, CLID UInt32, YCLID UInt64, ShareService String, ShareURL String, ShareTitle String, ParsedParams Nested(Key1 String, Key2 String, Key3 String, Key4 String, Key5 String, ValueDouble Float64), IslandID FixedString(16), RequestNum UInt32, RequestTry UInt8) ENGINE = MergeTree() PARTITION BY toYYYYMM(EventDate) ORDER BY (CounterID, EventDate, intHash32(UserID)) SAMPLE BY intHash32(UserID) SETTINGS index_granularity = 8192, storage_policy='s3_cache'"
+    # Avoid locking filesystem cache
+    clickhouse-client "SYSTEM STOP MERGES test.hits_s3"
     # AWS S3 is very inefficient, so increase memory even further:
     clickhouse-client --max_execution_time 600 --max_memory_usage 30G --max_memory_usage_for_user 30G --query "INSERT INTO test.hits_s3 SELECT * FROM test.hits SETTINGS enable_filesystem_cache_on_write_operations=0, max_insert_threads=16"
 

tests/queries/0_stateless/02240_system_filesystem_cache_table.sh

@@ -6,10 +6,12 @@
 CUR_DIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
 # shellcheck source=../shell_config.sh
 . "$CUR_DIR"/../shell_config.sh
+# shellcheck source=./cache.lib
+. "$CUR_DIR"/cache.lib
 
 for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
     echo "Using storage policy: $STORAGE_POLICY"
-    ${CLICKHOUSE_CLIENT} --query "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
     ${CLICKHOUSE_CLIENT} --query "SYSTEM DROP MARK CACHE"
     ${CLICKHOUSE_CLIENT} --query "SELECT count() FROM system.filesystem_cache"
 
@@ -30,7 +32,7 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
     ${CLICKHOUSE_CLIENT} --query "SELECT state, file_segment_range_begin, file_segment_range_end, size FROM system.filesystem_cache ORDER BY file_segment_range_begin, file_segment_range_end, size"
     ${CLICKHOUSE_CLIENT} --query "SELECT uniqExact(key) FROM system.filesystem_cache";
 
-    ${CLICKHOUSE_CLIENT} --query "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
     echo 'Expect no cache'
     ${CLICKHOUSE_CLIENT} --query "SELECT file_segment_range_begin, file_segment_range_end, size FROM system.filesystem_cache"
 
@@ -40,7 +42,7 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
     ${CLICKHOUSE_CLIENT} --query "SELECT state, file_segment_range_begin, file_segment_range_end, size FROM system.filesystem_cache ORDER BY file_segment_range_begin, file_segment_range_end, size"
     ${CLICKHOUSE_CLIENT} --query "SELECT uniqExact(key) FROM system.filesystem_cache";
 
-    ${CLICKHOUSE_CLIENT} --query "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
     echo 'Expect no cache'
     ${CLICKHOUSE_CLIENT} --query "SELECT file_segment_range_begin, file_segment_range_end, size FROM system.filesystem_cache"
 

tests/queries/0_stateless/02241_filesystem_cache_on_write_operations.sh

@@ -7,6 +7,8 @@
 CUR_DIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
 # shellcheck source=../shell_config.sh
 . "$CUR_DIR"/../shell_config.sh
+# shellcheck source=./cache.lib
+. "$CUR_DIR"/cache.lib
 
 for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
     echo "Using storage policy: $STORAGE_POLICY"
@@ -15,7 +17,8 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
     $CLICKHOUSE_CLIENT --echo --query "CREATE TABLE test_02241 (key UInt32, value String) Engine=MergeTree() ORDER BY key SETTINGS storage_policy='$STORAGE_POLICY', min_bytes_for_wide_part = 10485760, compress_marks=false, compress_primary_key=false, min_bytes_for_full_part_storage=0, ratio_of_defaults_for_sparse_serialization = 1"
     $CLICKHOUSE_CLIENT --echo --query "SYSTEM STOP MERGES test_02241"
 
-    $CLICKHOUSE_CLIENT --echo --query "SYSTEM DROP FILESYSTEM CACHE"
+    echo "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
 
     $CLICKHOUSE_CLIENT --echo --query "SELECT file_segment_range_begin, file_segment_range_end, size, state
     FROM
@@ -67,7 +70,8 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
 
     $CLICKHOUSE_CLIENT --echo --query "SELECT count(), sum(size) size FROM system.filesystem_cache"
 
-    $CLICKHOUSE_CLIENT --echo --query "SYSTEM DROP FILESYSTEM CACHE"
+    echo "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
 
     $CLICKHOUSE_CLIENT --echo --enable_filesystem_cache_on_write_operations=1 --query "INSERT INTO test_02241 SELECT number, toString(number) FROM numbers(100, 200)"
 

tests/queries/0_stateless/02286_drop_filesystem_cache.sh

@@ -1,11 +1,14 @@
 #!/usr/bin/env bash
-# Tags: long, no-fasttest, no-parallel, no-object-storage, no-random-settings
+# Tags: long, no-fasttest, no-parallel, no-object-storage, no-random-settings, no-flaky-check
+# Tag no-flaky-check -- access to system.remote_data_path is too slow with thread fuzzer enabled
 
 # set -x
 
 CUR_DIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
 # shellcheck source=../shell_config.sh
 . "$CUR_DIR"/../shell_config.sh
+# shellcheck source=./cache.lib
+. "$CUR_DIR"/cache.lib
 
 for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
     echo "Using storage policy: $STORAGE_POLICY"
@@ -17,15 +20,15 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
                                    SETTINGS storage_policy='$STORAGE_POLICY', min_bytes_for_wide_part = 10485760"
 
     $CLICKHOUSE_CLIENT --query "SYSTEM STOP MERGES test_02286"
-    $CLICKHOUSE_CLIENT --query "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
 
     $CLICKHOUSE_CLIENT --query "SELECT count() FROM system.filesystem_cache"
     $CLICKHOUSE_CLIENT --enable_filesystem_cache_on_write_operations=0 --query "INSERT INTO test_02286 SELECT number, toString(number) FROM numbers(100)"
 
     $CLICKHOUSE_CLIENT --query "SELECT * FROM test_02286 FORMAT Null"
     $CLICKHOUSE_CLIENT --query "SELECT count() FROM system.filesystem_cache"
 
-    $CLICKHOUSE_CLIENT --query "SYSTEM DROP FILESYSTEM CACHE"
+    drop_filesystem_cache
     $CLICKHOUSE_CLIENT --query "SELECT count() FROM system.filesystem_cache"
 
     $CLICKHOUSE_CLIENT --query "SELECT * FROM test_02286 FORMAT Null"
@@ -37,7 +40,6 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
 
     $CLICKHOUSE_CLIENT --query "SELECT * FROM test_02286 FORMAT Null"
     $CLICKHOUSE_CLIENT --query "SELECT count() FROM system.filesystem_cache"
-
     $CLICKHOUSE_CLIENT --query "SELECT count()
                                    FROM (
                                        SELECT
@@ -51,20 +53,26 @@ for STORAGE_POLICY in 's3_cache' 'local_cache' 'azure_cache'; do
                                    ON data_paths.cache_path = caches.cache_path"
 
     $CLICKHOUSE_CLIENT --query "DROP TABLE test_02286 SYNC"
-    $CLICKHOUSE_CLIENT --query "SELECT count() FROM system.filesystem_cache"
 
-    $CLICKHOUSE_CLIENT --query "SELECT cache_path FROM system.filesystem_cache"
-    $CLICKHOUSE_CLIENT --query "SELECT cache_path, local_path
-                                   FROM (
-                                       SELECT
-                                           arrayJoin(cache_paths) AS cache_path,
-                                           local_path,
-                                           remote_path
-                                       FROM
-                                           system.remote_data_paths
-                                       ) AS data_paths
-                                   INNER JOIN system.filesystem_cache AS caches
-                                   ON data_paths.cache_path = caches.cache_path"
+    cache_entries=$($CLICKHOUSE_CLIENT --query "SELECT count() FROM system.filesystem_cache")
+    echo "$cache_entries"
+    # system.remote_data_paths is very slow for web disks, so let's avoid extra
+    # call to it (we need it only for debugging of this tests, and only when we
+    # have cache entries, which tests does not expect)
+    if [ $cache_entries -gt 0 ]; then
+        $CLICKHOUSE_CLIENT --query "SELECT cache_path FROM system.filesystem_cache"
+        $CLICKHOUSE_CLIENT --query "SELECT cache_path, local_path
+                                       FROM (
+                                           SELECT
+                                               arrayJoin(cache_paths) AS cache_path,
+                                               local_path,
+                                               remote_path
+                                           FROM
+                                               system.remote_data_paths
+                                           ) AS data_paths
+                                       INNER JOIN system.filesystem_cache AS caches
+                                       ON data_paths.cache_path = caches.cache_path "
+    fi
 
     $CLICKHOUSE_CLIENT --query "DROP TABLE IF EXISTS test_022862"
 done

tests/queries/0_stateless/cache.lib

@@ -0,0 +1,9 @@
+#!/usr/bin/env bash
+
+function drop_filesystem_cache()
+{
+    while [ $($CLICKHOUSE_CURL -sS "$CLICKHOUSE_URL" -d "SELECT count() FROM system.filesystem_cache") -gt 0 ]; do
+        $CLICKHOUSE_CURL -sS "$CLICKHOUSE_URL" -d "SYSTEM DROP FILESYSTEM CACHE"
+        sleep 1
+    done
+}