Bug 2041570 - Lost all collected metrics data when both Prometheus pods restarted
Summary: Lost all collected metrics data when both Prometheus pods restarted
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: Monitoring
Version: 4.10
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
: ---
Assignee: Brad Ison
QA Contact: Junqi Zhao
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2022-01-17 17:56 UTC by Denys Shchedrivyi
Modified: 2022-01-18 10:37 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2022-01-18 10:37:50 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)

Description Denys Shchedrivyi 2022-01-17 17:56:24 UTC
Description of problem:
 After draining nodes both pods (prometheus-k8s-0 and prometheus-k8s-1) restarted and all data that was previously collected by metrics removed from a cluster


Version-Release number of selected component (if applicable):
CNV-4.10

How reproducible:
100%

Steps to Reproduce:
1. drain first node where pod prometheus-k8s-0 is running
2. return node back and drain second node where prometheus-k8s-1 is running
3. check metrics on a cluster - old data removed

Actual results:
 Metrics cleared when both pods prometheus-k8s-0 and prometheus-k8s-1 restarted

Expected results:
 Old data should be kept 

Additional info:
 I guess some persistent storage should be using for keeping all metrics:
https://docs.openshift.com/container-platform/4.9/monitoring/configuring-the-monitoring-stack.html#configuring-persistent-storage

Comment 1 Brad Ison 2022-01-18 10:37:50 UTC
If you haven't configured persistent storage, this is expected. Following the procedure in the docs you linked is a prerequisite to having the data persist.


Note You need to log in before you can comment on or make changes to this bug.