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鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Search result count widget not caching #1640

Closed
boosty opened this Issue Dec 17, 2015 · 7 comments

Comments

Projects
None yet
4 participants
@boosty

boosty commented Dec 17, 2015

First of all I want to say thank you for Graylog! You know that it took us a while to set it up properly for our use case at XING, but we now reached a pretty stable state 馃槃

We found one small issue when using search result count widgets: The number seems to get recalculated every minute or so, even though the cache time has been set to 3600s.

We see this not only by the number frequently changing, but also the corresponding metric ("calculations") increasing quickly.

Any ideas?

screen shot 2015-12-17 at 18 25 59
screen shot 2015-12-17 at 18 26 27

@boosty

This comment has been minimized.

boosty commented Dec 17, 2015

Btw.: We are using 1.3.0.

@jfmarquet

This comment has been minimized.

jfmarquet commented Jan 8, 2016

Hello,

I have the same issue with version 1.3.0, the widget cache time is configured to 3600 but it seems like there is no cache at all. The values are refreshed every time they are requested.
The same configuration was working fine with version 1.2.2. Do you have any update on this topic?
Thanks

@kroepke

This comment has been minimized.

Member

kroepke commented Jan 8, 2016

This should not have changed and is a bit weird, since all widgets run through the same cache.
However the cache is per server process, so unless you happen to hit the same server on each request, it will have to recalculate it up to the number of servers you are running.

We do not have cluster-wide caches yet.

@jfmarquet

This comment has been minimized.

jfmarquet commented Jan 8, 2016

Hello,
Thanks for your reply. The current configuration is running on a single machine with only one graylog-server process so I don't think we face a cluster issue here. Would there be anything to check on our side, any log or info we could provide you to help finding the issue?
Thank you.

@kroepke

This comment has been minimized.

Member

kroepke commented Jan 8, 2016

In fact we have just determined that this is a bug introduced in 1.3, we will be working on a fix for this problem in a bugfix release.

@joschi joschi self-assigned this Jan 8, 2016

@joschi joschi added this to the 1.3.3 milestone Jan 8, 2016

joschi added a commit that referenced this issue Jan 8, 2016

Revert removal of DashboardRegistry
This reverts commit 8336a8d and f997da6.
Fixes #1640, refs #1588, refs #1542
@boosty

This comment has been minimized.

boosty commented Jan 10, 2016

Cool, thanks a lot!

@joschi

This comment has been minimized.

Contributor

joschi commented Jan 11, 2016

Fixed by merging #1681.

@joschi joschi closed this Jan 11, 2016

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment