Note: As a recurring issue, this issue should always be open.
TODO(mattjmcnaughton) Determine how to track/organize recurring issues.
Issue
At a regular cadence, I want to ensure that my k8s cluster is making optimal usage of resources. In other words, I want to be paying as little as possible.
Impl
The first time I do this, I should create a check list. Possible ideas include...
Pods which can have their resource limits decreased (and as a result, decrease in the number of workers)
PVCs which can be decreased.
Convert EC2 instances from on-demand to spot or reserved.
Decrease EBS volumes mounted to nodes.
Merge services with individual ELBs onto shared ingress
mattjmcnaughton
commented
5 years ago
Note: As a recurring issue, this issue should always be open.
TODO(mattjmcnaughton) Determine how to track/organize recurring issues.
Issue At a regular cadence, I want to ensure that my k8s cluster is making optimal usage of resources. In other words, I want to be paying as little as possible.
Impl The first time I do this, I should create a check list. Possible ideas include...