-
Task
-
Resolution: Done
-
Medium
-
None
-
None
-
None
Summary: An issue has been detected in the Helm charts templates for OOM where specified resource limits and requests were not being applied due to an indentation error. As a result, the service may have been utilizing more or fewer resources than intended, potentially impacting its performance and stability. This ticket aims to rectify this issue and introduce new default values for resource allocation.
All proposed limits and request were tested and recommended by Goldilocks :
https://www.fairwinds.com/blog/introducing-goldilocks-a-tool-for-recommending-resource-requests
Also all CPU LIMITS are unlimited because of we want to prevent CPU throttling and insufficient CPU without limits
You can also read this: https://home.robusta.dev/blog/stop-using-cpu-limits
Description:
This ticket relates to the misconfiguration of Kubernetes resource limits and requests in the OOM helm charts. We’ve detected that due to an indentation error in the YAML file, the limits and requests for the CPU and memory resources weren’t being correctly applied as fixed in this ticket https://jira.onap.org/browse/OOM-3104.
This ticket is fixing this OOM helm charts components:
aai
cds
cps
dcaegen2-services
dmaap
holmes
msb
nbi
oof
platform
policy
sdc
sdnc
so