Skip to content
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

Incorrect calculation for GC Count in Stack Monitoring #31191

Open
cachedout opened this issue Feb 14, 2019 · 5 comments
Open

Incorrect calculation for GC Count in Stack Monitoring #31191

cachedout opened this issue Feb 14, 2019 · 5 comments
Labels
Team:Monitoring Stack Monitoring team

Comments

@cachedout
Copy link
Contributor

Kibana version:
8c8a4a2

Elasticsearch version:
7.0 snapshot

Server OS version:
OS X

Browser version:
Version 71.0.3578.98 (Official Build) (64-bit)

Browser OS version:
OS X

Original install method (e.g. download page, yum, from source, etc.):
Source

Describe the bug:

On the GC Count graph that exists on an Elasticsearch node page inside Stack Monitoring, it appears that the metric is either being calculated incorrectly or the metric explanation and label needs to be changed.

What it appears that this metric is describing is the number of garbage collection events which have occurred during a thirty second period. However, in the graph linked below, you'll see that the number is listed as hovering between 0.1 and 0.05 events. This is impossible. You can't have a tenth of an event. :)

It does not appear that this is problem on the Elasticsearch side as a brief sanity check reveals that the metric appears to be a counter which increments upward correctly.

Steps to reproduce:

  1. Create an Elaticsearch cluster which is relatively idle and enable Stack Monitoring for it.
  2. Navigate to an Elasticsearch Node page inside of Stack Monitoring
  3. Observe that the values in the GC Count may be < 1.

Expected behavior:

Screenshots (if relevant):
screen shot 2019-02-14 at 12 10 16
Errors in browser console (if relevant):

Provide logs and/or server output (if relevant):

Any additional context:

@cachedout cachedout added the Team:Monitoring Stack Monitoring team label Feb 14, 2019
@elasticmachine
Copy link
Contributor

Pinging @elastic/stack-monitoring

@cachedout
Copy link
Contributor Author

cc: @pickypg for historical context to see if we can learn more about the original intention here.

@pickypg
Copy link
Member

pickypg commented Feb 14, 2019

@tsullivan Do you remember what version we switched to /s derivative normalization?

EDIT: This is why you see 0.1 because they're 30 second buckets and there must have been 3 young GCs in the bucket.

@pickypg
Copy link
Member

pickypg commented Feb 14, 2019

I think that we probably need to update the labels for anything using a normalized derivative.

@tsullivan
Copy link
Member

@tsullivan Do you remember what version we switched to /s derivative normalization?

In the original repo, that PR was labeled for 6.3.0/7.0.0

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Team:Monitoring Stack Monitoring team
Projects
None yet
Development

No branches or pull requests

4 participants