[jboss-jira] [JBoss JIRA] (WFLY-5655) Clustering performance regression in read-heavy scenarios

[Michal Vinkler (JIRA)]

Michal Vinkler created WFLY-5655:
------------------------------------

             Summary: Clustering performance regression in read-heavy scenarios
                 Key: WFLY-5655
                 URL: https://issues.jboss.org/browse/WFLY-5655
             Project: WildFly
          Issue Type: Bug
          Components: Clustering
            Reporter: Michal Vinkler
            Assignee: Paul Ferraro
            Priority: Blocker


Our performance tests show regression in read-heavy scenarios with the default settings. Even with the JBEAP-1740 performance fix applied, the test progress is very different from the EAP6 tests.

Links: scenario heavyread-session-dist-sync
[EAP 6.4.0.GA|http://jenkins.mw.lab.eng.bos.redhat.com/hudson/job/eap-6x-stress-heavyread-session-dist-sync_noperf21/1/artifact/report/graph-throughput.png]
[EAP7 DR13|http://jenkins.mw.lab.eng.bos.redhat.com/hudson/job/eap-7x-stress-heavyread-session-dist-sync/2/artifact/report/graph-throughput.png]
[EAP7 DR13 with applied JBEAP-1740|http://jenkins.mw.lab.eng.bos.redhat.com/hudson/job/eap-7x-stress-heavyread-session-dist-sync-bigger-oob-threadpool/1/artifact/report/graph-throughput.png]

While changing max-unreplicated-interval to 0 in EAP 6 tests may change the test progress to be similar to the EAP 7 tests, performance results with EAP7 using the *default settings* have to be comparable with the results from EAP6. 




--
This message was sent by Atlassian JIRA
(v6.4.11#64026)