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

[Monitoring] Dot prefixed indices are ignored for Alerts #99396

Closed
simianhacker opened this issue May 5, 2021 · 6 comments · Fixed by #100831
Closed

[Monitoring] Dot prefixed indices are ignored for Alerts #99396

simianhacker opened this issue May 5, 2021 · 6 comments · Fixed by #100831
Assignees

Comments

@simianhacker
Copy link
Member

Kibana version:

master

Elasticsearch version:

master

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

source

Describe the bug:

I was testing alerting with some data from Agent, specifically the "Shard size" alert. I set the alert to a ridiculously low value to ensure it would still fire. When it didn't fire I started to investigate and realized that we are ignoring dot prefixed indices. This seems like a huge hole in our alerting since dot prefixed indices can grow as well.

Steps to reproduce:

  1. Index data from Elastic Agent
  2. Edit the "Shard size" alert (via setup mode on the overview page) and set the size to 0.00000000000001
  3. The alert will never fire

Expected behavior:

I would expect that alerts are run on all indices regardless of the name when the index pattern is set to *. I understand that dot prefix indices are technically "hidden" but they are also resource that can grow and have issues.

@simianhacker simianhacker added bug Fixes for quality problems that affect the customer experience Feature:Stack Monitoring labels May 5, 2021
@botelastic botelastic bot added the needs-team Issues missing a team label label May 5, 2021
@brianseeders brianseeders added the Team:Monitoring Stack Monitoring team label May 18, 2021
@elasticmachine
Copy link
Contributor

Pinging @elastic/stack-monitoring (Team:Monitoring)

@botelastic botelastic bot removed the needs-team Issues missing a team label label May 18, 2021
@igoristic
Copy link
Contributor

@simianhacker This was done intentionally: #89410 (comment), but you have a valid point.

Maybe a better UX would be to default to something like *,-.*, so there's still the option to include internal indices if needed.

cc: @ravikesarwani

@ravikesarwani
Copy link
Contributor

In my mind alerts needs to be actionable by the users. Indices starting with "." are internal indices that we as Elastic own and not something that the solution users should really care about. If we are really growing our internal indices to more than the recommended size (50 GB, but we are using 55 GB default to account for 10% grace) then we should fix that rather than alert users about it.

@igoristic
Copy link
Contributor

@ravikesarwani I see your point, but the user can still control that by proxy, eg: to control .monitoring-* index size they can change: xpack.monitoring.collection.interval and xpack.monitoring.history.duration. Other apps (that are managing internal indices), will probably have similar settings/options

@igoristic
Copy link
Contributor

igoristic commented May 24, 2021

@ravikesarwani @simianhacker I'm going to add the *,-.* default. Let me know if there are any objections

@igoristic igoristic removed the bug Fixes for quality problems that affect the customer experience label May 24, 2021
@ravikesarwani
Copy link
Contributor

I am good with that. It makes explicit that we are excluding dot prefixed indices by default.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging a pull request may close this issue.

5 participants