[
https://issues.apache.org/jira/browse/IGNITE-9710?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16651305#comment-16651305
]
Alexey Goncharuk commented on IGNITE-9710:
------------------------------------------
[~andrey-kuznetsov], in my opinion, such approach to caches start defeats the
purpose of the watchdog service. All potentially dangerous operations are now
wrapped in {{blockingSectionBegin/blockingSectionEnd}}. Instead, we should
insert thread progress markers to all loops where we iterate over caches (for
example, inside {{forAllRegisteredCacheGroups}}) and over partitions. This way
we will still report a long-running disk IO operations, but will tolerate a
large number of starting caches.
> Ignite watchdog service handles longrunning cache creation
> ----------------------------------------------------------
>
> Key: IGNITE-9710
> URL: https://issues.apache.org/jira/browse/IGNITE-9710
> Project: Ignite
> Issue Type: Bug
> Components: cache
> Affects Versions: 2.6
> Reporter: Vyacheslav Daradur
> Assignee: Andrey Kuznetsov
> Priority: Major
> Fix For: 2.7
>
> Attachments: LongRunningCacheCreationTest.java
>
>
> Ignite watchdog service introduced by IGNITE-6587 handles long running cache
> creation.
> Action in {{GridDhtPartitionsExchangeFuture#init}} may take significant time
> and possibly should be covered by blocking section of warchdog service.
> Reproducer was attached: [^LongRunningCacheCreationTest.java].
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
