-
Notifications
You must be signed in to change notification settings - Fork 8.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Constantly memory leaking on x-pack.reporting #90274
Comments
@drTr0jan To help us narrow down potential causes, what are you using Kibana for? Your memory graph doesn't show a lot of variation within a day like what I would expect if you had a lot of users using Kibana during work hours and then dropping again when users sleep. |
@rudolf, generally I'm using Kibana only for viewing syslogs (Elasticsearch indices). But the Kibana from a screenshot graph is an instance that users don't working (using) with. |
We have not had any other reports of Kibana 7.10.2 leaking memory, so I suspect this is FreeBSD related. As far as I know there's nothing in the x-pack.ml plugin that should be platform-specific. However, the reporting plugin uses chrome to render reports so I think it's the most likely culprit. Can you try with only x-pack.reporting disabled? (since FreeBSD isn't a supported platform the support we're able to provide is limited, see https://www.elastic.co/support/matrix) |
Pinging @elastic/kibana-core (Team:Core) |
@rudolf, I've enabled x-pack.ml plugin two days ago. Kibana is working correctly. |
Pinging @elastic/kibana-reporting-services (Team:Reporting Services) |
Pinging @elastic/kibana-app-services (Team:AppServices) |
@elastic/kibana-app-services: If it is reporting and reporting is not really used then the code in question (task manager) will be removed in the next release (after 7.12) |
The code in question here is ESQueue, which uses a loop to poll for Reporting jobs to claim. If the Reporting feature is enabled but reports aren't generated, the loop still runs. After 7.12, we are removing ESQueue and replacing it with a Kibana-wide service called Task Manager. |
I missed this the first time, but it should be noted that FreeBSD is not listed as a supported OS in the Kibana support matrix: https://www.elastic.co/support/matrix Also, if you are not using ML or Reporting then I don't see how they can be responsible for a memory leak. My guess is that disabling those plugins allowed less code. Node seemed to have more success running under a 1GB memory restriction. Closing as not reproducible / not support OS. |
@drTr0jan Have you found any solution for that yet? I'm running into the same problem on freebsd... |
@L4rS6 try to disable x-pack.reporting. See Bug 253314 - textproc/kibana7: fix memleak and some port improvements for more info. |
@drTr0jan Thank you, that solved the problem. |
--max_old_space_size=1024
)I've two installation of Kibana with same config and each of them has constantly memory leaking about ~100 MB/day.
After node has reached about 1163 MB of resident set memory consumption (by
ps(1)
) then it falls with fatal error -JavaScript heap out of memory
.The text was updated successfully, but these errors were encountered: