[
https://issues.apache.org/jira/browse/IGNITE-12119?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16918462#comment-16918462
]
Ignite TC Bot commented on IGNITE-12119:
----------------------------------------
{panel:title=Branch: [pull/6821/head] Base: [master] : No blockers
found!|borderStyle=dashed|borderColor=#ccc|titleBGColor=#D6F7C1}{panel}
[TeamCity *--> Run :: All*
Results|https://ci.ignite.apache.org/viewLog.html?buildId=4545846&buildTypeId=IgniteTests24Java8_RunAll]
> Peer Class Loading has no retries
> ---------------------------------
>
> Key: IGNITE-12119
> URL: https://issues.apache.org/jira/browse/IGNITE-12119
> Project: Ignite
> Issue Type: Bug
> Reporter: Ivan Bessonov
> Assignee: Ivan Bessonov
> Priority: Major
> Fix For: 2.8
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> That's it. Peer Class Loading has short timeout and no retries, and if it
> fails, loading of class will not be reattempted.
> I believe this is in part because GridDeploymentClassLoader is a class
> loader. If it throws ClassNotFoundException when asked to load class, JVM
> will take notice and not reattempt to load this class, even if error was
> transient.
> Proposed amendments:
> * Increase timeouts, introduce immediate retries.
> * See if we can report transient class loading issue to JVM.
> * If all failed, we need to mark class loader as invalid when timeout
> occurs, phase out its usage and create a new class loader which will
> reattempt to load this class later.
> Please note that extended waiting in class loader is not recommended because
> it can cause grid to stall.
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
