[
https://issues.apache.org/jira/browse/IGNITE-23399?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17889160#comment-17889160
]
Kirill Tkalenko commented on IGNITE-23399:
------------------------------------------
Looks good.
> Decrease frequency of Metastorage idle SafeTime propagation in tests
> --------------------------------------------------------------------
>
> Key: IGNITE-23399
> URL: https://issues.apache.org/jira/browse/IGNITE-23399
> Project: Ignite
> Issue Type: Improvement
> Reporter: Roman Puchkovskiy
> Assignee: Roman Puchkovskiy
> Priority: Major
> Labels: ignite-3
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> It's currently 10ms which seems too little, 50ms should be fine (as
> DelayDuration in tests is 100ms).
> The value of 10ms means that we write up to 100 times per second per node to
> the Metastorage log (and each write causes an fsync) which might create heavy
> (and unnecessary) load. This might be the cause of some integration tests
> (which start 5-6 Ignite nodes) failing at TC from time to time.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
