[FLINK-26431] Release Testing: Exception history for the adaptive scheduler

• Type: Improvement
• Status: Resolved
• Priority: Blocker
• Resolution: Done
• Affects Version/s: 1.15.0
• Fix Version/s: 1.15.0
• Component/s: Runtime / Coordination
• Labels: release-testing
• Assignee: Liu
• Reporter: Niklas Semmler
• Created: 01/Mar/22 14:00
• Updated: 03/Mar/22 16:12
• Resolved: 03/Mar/22 16:12

Description

FLINK-21439 adds support for the exception history to the adaptive scheduler.

To evaluate the functionality follow the following steps:

Add the following lines to your flink config

taskmanager.host: localhost # a workaround
scheduler-mode: reactive
restart-strategy: fixeddelay
restart-strategy.fixed-delay.attempts: 100

Create the test job JAR and move it to the usrlib folder. (You may have to create it.) Here "$FLINK-HOME" refers to the folder of your flink installation.

$ git clone [e-mail]:metaswirl/checkpoint-crasher-job.git
$ cd checkpoint-crasher-job
$ mvn install
$ mv target/checkpoint-crasher-job-0.1.jar $FLINK_HOME/usrlib

Start the job

$ bin/taskmanager.sh start
$ bin/taskmanager.sh start
$ bin/taskmanager.sh start
$ bin/standalone-job.sh start --main-class org.apache.niklassemmler.flinkjobs.CheckpointCrasherJob

You should be able to find the exception `java.lang.RuntimeException: Test exception` in the web interface under Jobs->Running Jobs->Exceptions.

Issue Links

• is caused by: FLINK-21439 Adaptive Scheduler: Add support for exception history