-
Bug
-
Resolution: Done-Errata
-
Normal
-
4.19
-
None
-
Quality / Stability / Reliability
-
False
-
-
None
-
None
-
None
-
None
-
None
-
None
-
Done
-
None
-
N/A
-
None
-
None
-
None
-
None
Description of problem:
At present, the way of gather.json to print some error logs is as below:
["unexpected end of JSON input","unexpected end of JSON input","unexpected end of JSON input"]
It's OK in most cases, but for the cluster that has more worker nodes, for example, if in a cluster has 100 worker nodes, this file is going to include 100 duplicate error logs, so we should organize these logs in a better way.
Version-Release number of selected component (if applicable):
How reproducible:
1. Update the IO image to the debugging image, so IO can request the target pods instead of the true extractor pods, which can mock the abnormal scenarios.
2. Create several pods to mock extractor pods to return responses without the right JSON format.
3. Restarting IO.
4. View the content of gather.json in the Insights archive.
Steps to Reproduce:
1. Update the IO image to the debugging image, so IO can request the target pods instead of the true extractor pods, which can mock the abnormal scenarios. 2. Create several pods to mock extractor pods to return responses without the right JSON format. 3. Restarting IO. 4. View the content of gather.json in the Insights archive.
Actual results:
4. gather.json includes some duplicate error logs
Expected results:
4. Should optimize printing method to display the error logs better.
Additional info:
Suggestion: Will it make more sense to print the relevant node IP of the error log?.
- clones
-
-
- Verified
-
- depends on
-
-
- Verified
-
- links to
-
RHBA-2025:10771
OpenShift Container Platform 4.19.4 bug fix update
- mentioned on
