[
https://issues.apache.org/jira/browse/IGNITE-23532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18055935#comment-18055935
]
Filipp Shergalis edited comment on IGNITE-23532 at 2/2/26 2:15 PM:
-------------------------------------------------------------------
Throttling was introduced
was (Author: JIRAUSER303902):
Throttling
> Fix checkpoint timeout under high load
> --------------------------------------
>
> Key: IGNITE-23532
> URL: https://issues.apache.org/jira/browse/IGNITE-23532
> Project: Ignite
> Issue Type: Improvement
> Components: persistence
> Reporter: Filipp Shergalis
> Priority: Major
> Labels: ignite-3
>
> See comment in IGNITE-23325 for detailed explanation
>
> Not reproducible in main right now, but is a problem i.e. with optimizations
> in branch *ignite-22835-2 -* writes happen so fast, that we're bound by
> disk's fsync speed
>
> When writing to pages too fast, we have to schedule a new checkpoint while
> the last one is still in progress, and timeout a write if 10 seconds (by
> default) passed before next checkpoint starts
>
> Proposed solution is to port throttling from Ignite 2
--
This message was sent by Atlassian Jira
(v8.20.10#820010)