Hi, we use the flink k8s operator. Just want to ch...
# troubleshooting
f
Hi, we use the flink k8s operator. Just want to check, in
FlinkDeployments
is it by design that task manager pods always have limits set on them? Is there no way to opt out? The podTemplate I have is not respected. the cpu.limit is always set as cpu.request. We saw a configuration
cpu.limit-factor=1
. If one wants to opt out of limits, should I just set this to ♾️ ?
g
You could set the limit factor to a large number. No limit does not make much operational sense
✅ 1
y
in 1.5 operator
kubernetes.taskmanager.cpu.limit-factor=N
has no effect. I couldn't see this setting in source code (branch 1.5 or main)