Chi-Hsuan Huang created HDDS-17812:
--------------------------------------

             Summary: Speed up integration tests waiting for closed pipelines 
to be scrubbed
                 Key: HDDS-17812
                 URL: https://issues.apache.org/jira/browse/HDDS-17812
             Project: Apache Ozone
          Issue Type: Improvement
          Components: test
            Reporter: Chi-Hsuan Huang


Since HDDS-9479, a closed pipeline stays in CLOSED state until the scrubber 
deletes it after {{ozone.scm.pipeline.destroy.timeout}}. Tests that wait for 
the pipeline to go away got the same config block: close container wait 2s, 
scrub interval 2s, destroy timeout 5s. This block is now in 11 integration test 
classes.

Each such wait takes up to about 7 seconds. For example, every decommission or 
maintenance in {{TestDecommissionAndMaintenance}} waits for the node's 
pipelines to be scrubbed before moving on.

Changing the block to 1s / 1s / 2s keeps the destroy timeout longer than the 
close container wait, as HDDS-9479 intended. Local run time:

|| ||Before||After||
|TestDecommissionAndMaintenance|273-284s|238s|
|TestReplicationManagerIntegration|166-172s|141s|
|TestContainerStateMachineFailures|96-104s|91s|




--
This message was sent by Atlassian Jira
(v10.3.26#10030026)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to