[
https://issues.apache.org/jira/browse/TS-4328?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15257595#comment-15257595
]
ASF GitHub Bot commented on TS-4328:
------------------------------------
Github user jacksontj commented on the pull request:
https://github.com/apache/trafficserver/pull/554#issuecomment-214621535
Well, I'm not sure that I agree it violates the encapsulation-- since its
effectively a variable that tracks timings (which are only set in specific
states in the state machine-- IIRC most are only set once) which makes it more
or less a history of where the HTTPSM has gone. For this particular feature I
need to know "have we sent bytes" and today the only way that is stored is in
the milestones. It sounds like both of you guys ( @zwoop and @jpeach ) want me
to add a bool to the state machine to store this-- which I can do, but since
the milestone is only set in one place, it'll effectively be the same check--
so i still don't really get how that would be "cleaner" in any way :/
> Connect_retries re-connects even if request made it to origin (TS-3440 repeat)
> ------------------------------------------------------------------------------
>
> Key: TS-4328
> URL: https://issues.apache.org/jira/browse/TS-4328
> Project: Traffic Server
> Issue Type: Bug
> Reporter: Thomas Jackson
> Assignee: Thomas Jackson
> Fix For: sometime
>
>
> I'm creating another issue as the root bug for TS-3440 can still be
> reproduced if the origin is extremely slow (but no returns no error response).
> I've opened a PR (https://github.com/apache/trafficserver/pull/554) which
> reverts the original patch and instead checks if any bytes where sent.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
