[
https://issues.apache.org/jira/browse/AMQ-7131?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16745230#comment-16745230
]
Timothy Bish commented on AMQ-7131:
-----------------------------------
[~graben] while the change looks a bit odd it isn't much worse than some of the
other code in there that works around behaviours in the commons pool that clash
with how a Connection pool needs to operate so that isn't as bad as one might
think. The connect timeout fix seems more awkward to me because it forces a
lengthy wait on something we can know is going to fail and then puts the
requirement on the caller to respond appropriately to that failure.
It doesn't appear there is anything else we can do in the pool configuration to
prevent the key from being removed when the entries hit zero due to the single
connection being evicted but maybe you want to poke around in there to see for
sure.
> ActiveMQ JMS pool has no borrow timeout causing starvation
> ----------------------------------------------------------
>
> Key: AMQ-7131
> URL: https://issues.apache.org/jira/browse/AMQ-7131
> Project: ActiveMQ
> Issue Type: Bug
> Components: activemq-pool
> Affects Versions: 5.15.8
> Reporter: Benjamin Graf
> Assignee: Benjamin Graf
> Priority: Major
>
> A maxPoolSize of 1 can cause starvation during createConnection if the only
> connection gets evicted concurrently since borrow method of pool has no
> timeout waiting forever.
> Maybe introduce a new variable connectionTimeout with default 30000ms.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
