[
https://issues.apache.org/jira/browse/TS-4507?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15347351#comment-15347351
]
ASF GitHub Bot commented on TS-4507:
------------------------------------
Github user shinrich commented on the issue:
https://github.com/apache/trafficserver/pull/717
Yes, this is snakey logic. I have a couple more iterations after working
with various teams internally. I'll be putting up a new PR for that
(really_destroy goes away), so I'm closing this PR for now. It will at a
minimum have a healthy comment discussion on the flow.
As far as testing goes, I will need some help with that. Ultimately, some
of that will need to be shaken out through load testing with appropriate
plugins. We have an intern writing up some functional test plugin archetypes
to help with that.
But I agree we should be able run through some basic cases in an
integration testing framework. I'll write up a list of test case requirements
and either work with @dragon512 to get this exercised in next gen TSQA or work
with someone else to get this going in current TSQA.
> It is still possible for SSN_CLOSE hook to be called before TXN_CLOSE hook
> --------------------------------------------------------------------------
>
> Key: TS-4507
> URL: https://issues.apache.org/jira/browse/TS-4507
> Project: Traffic Server
> Issue Type: Bug
> Components: Core
> Reporter: Susan Hinrichs
> Assignee: Susan Hinrichs
> Fix For: 7.0.0
>
>
> One of our plugins will occasionally crash. It appears there is still a path
> for HTTP2 that has the SSN_CLOSE hook close before the TXN_CLOSE hook.
> Working through solutions that delay the SSN_CLOSE hook until after all the
> TXN_CLOSE hooks, but does not lose the SSN_CLOSE.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
