[
https://issues.apache.org/jira/browse/MRESOLVER-347?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17705380#comment-17705380
]
ASF GitHub Bot commented on MRESOLVER-347:
------------------------------------------
mvitz commented on PR #274:
URL: https://github.com/apache/maven-resolver/pull/274#issuecomment-1485185661
@michael-o I started the discussion on the mailing list and asked if it's
feasible to add some kind of configuration to resolve the problem of failing
builds because the pool connection TTL is higher than the Azure limit.
From my perspective, there are four possible ways to solve it.
1. The user can deactivate the pooling altogether.
2. The user can specify the TTL.
3. The default TTL is set lower than Azure.
4. The HttpClient can detect this kind of stale connection.
Point 3. would, of course, be good because the next one will not run into
the problem altogether, but even then, it may be useful to be able to configure
this if in a specific environment the TTL can't be longer than e.g. two minutes.
Therefore for me point 2. would be a good enough solution maybe combined
with a more sensible default (= 3.) on top.
> Transport http should be able to not reuse or maximize TTL of connections
> -------------------------------------------------------------------------
>
> Key: MRESOLVER-347
> URL: https://issues.apache.org/jira/browse/MRESOLVER-347
> Project: Maven Resolver
> Issue Type: Improvement
> Components: Resolver
> Reporter: Tamas Cservenak
> Priority: Major
> Fix For: 1.9.8
>
>
> In some environments (like Azure) the connections should not be reused, or
> TTL adjusted, as Azure is killing them automagically.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
