Right now any deploy can use the amount of resources it needs, it is controlled during the code review phase. We should set up a cluster mechanism to protect the cluster from resource starvation, this is also a prerequisite for enabling autoscaling in the cluster.
Related Gerrit Patches:
|operations/deployment-charts : master||k8s: enable limitranges and resource quotas in staging|
|Open||None||T212123 Kubernetes clusters roadmap|
|Open||None||T228965 set up limitranges and resourcequotas to protect the cluster from resource abuse and starvation|