Description

<What were you trying to do that didn't work?>

When I created a kataconfig, I noticed that the controller-manager pod restarted 2 times due to OOMKILL.

Steps to reproduce

<What actions did you take to hit the bug?>
1. Install OSC
2.Create kataconfig with peerpods enabled
3. watch controller-manager pod

Expected result

<What did you expect to happen?>

controller-manager does not get OOMKILLed

Actual result

<What actually happened?>

Gets OOMKILLed

Impact

<How badly does this interfere with using the software?>

Loose information about what happened during the time killed instances were running.  Previous logs are not available.  No noticiable end user affects.

Env

<Where was the bug found, i.e. OCP build, operator build, kata-containers build, cluster infra, test case id>

OSC 1.6.0-41 OCP 4.15.9 and OCP 4.15.10

Additional helpful info

<logs, screenshot, doc links, etc.>

When the pod stays up and running, pod metrics show that it is running right at 100Mi for memory. memory limit is set at 100Mi and request if 40Mi.  I think these need to be updated.  I would suggest request of 100Mi and limit of 150Mi to be safe.