[FLINK-21431] UpsertKafkaTableITCase.testTemporalJoin hang

• Type: Bug
• Status: Reopened
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.12.2, 1.13.0
• Fix Version/s: 1.12.8, 1.13.7
• Component/s: Connectors / Kafka
• Labels: auto-deprioritized-critical auto-unassigned pull-request-available test-stability
• Assignee: Unassigned
• Reporter: Guowei Ma
• Created: 22/Feb/21 03:01
• Updated: 01/Feb/22 14:47

Description

This case hangs almost 3 hours:

https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=13543&view=logs&j=ce8f3cc3-c1ea-5281-f5eb-df9ebd24947f&t=f266c805-9429-58ed-2f9e-482e7b82f58b

Test testTemporalJoin[format = csv](org.apache.flink.streaming.connectors.kafka.table.UpsertKafkaTableITCase) is running. -------------------------------------------------------------------------------- 23:08:43,259 [ main] INFO org.apache.flink.streaming.connectors.kafka.KafkaTestEnvironmentImpl [] - Creating topic users_csv 23:08:45,303 [ main] WARN org.apache.flink.streaming.connectors.kafka.FlinkKafkaProducer [] - Property [transaction.timeout.ms] not specified. Setting it to 3600000 ms 23:08:45,430 [ChangelogNormalize(key=[user_id]) -> Calc(select=[user_id, user_name, region, CAST(modification_time) AS timestamp]) -> Sink: Sink(table=[default_catalog.default_database.users_csv], fields=[user_id, user_name, region, timestamp]) (1/1)#0] WARN org.apache.flink.streaming.connectors.kafka.FlinkKafkaProducer [] - Using AT_LEAST_ONCE semantic, but checkpointing is not enabled. Switching to NONE semantic. 23:08:45,438 [ChangelogNormalize(key=[user_id]) -> Calc(select=[user_id, user_name, region, CAST(modification_time) AS timestamp]) -> Sink: Sink(table=[default_catalog.default_database.users_csv], fields=[user_id, user_name, region, timestamp]) (1/1)#0] INFO org.apache.flink.streaming.connectors.kafka.FlinkKafkaProducer [] - Starting FlinkKafkaInternalProducer (1/1) to produce into default topic users_csv 23:08:45,791 [Source: TableSourceScan(table=[[default_catalog, default_database, users_csv, watermark=[CAST($3):TIMESTAMP(3)]]], fields=[user_id, user_name, region, timestamp]) (1/1)#0] INFO org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumerBase [] - Consumer subtask 0 has no restore state. 23:08:45,810 [Source: TableSourceScan(table=[[default_catalog, default_database, users_csv, watermark=[CAST($3):TIMESTAMP(3)]]], fields=[user_id, user_name, region, timestamp]) (1/1)#0] INFO org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumerBase [] - Consumer subtask 0 will start reading the following 2 partitions from the earliest offsets: [KafkaTopicPartition{topic='users_csv', partition=1}, KafkaTopicPartition{topic='users_csv', partition=0}] 23:08:45,825 [Legacy Source Thread - Source: TableSourceScan(table=[[default_catalog, default_database, users_csv, watermark=[CAST($3):TIMESTAMP(3)]]], fields=[user_id, user_name, region, timestamp]) (1/1)#0] INFO org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumerBase [] - Consumer subtask 0 creating fetcher with offsets {KafkaTopicPartition{topic='users_csv', partition=1}=-915623761775, KafkaTopicPartition{topic='users_csv', partition=0}=-915623761775}. ##[error]The operation was canceled.

Issue Links

• relates to: FLINK-21516 UpsertKafkaTableITCase.testTemporalJoin fails on AZP
• links to: GitHub Pull Request #15578