Illya Chekrygin
Illya Chekrygin
### Overview Currently, resource check runs in equal intervals with either result in: a) No Op b) Scale Up c) Scale Down  The resource utilization that resembles above pattern...
For larger clusters (100+ nodes), scaling down (say to 20 nodes) - takes an order of several hours due to a conservative nature of scale-down step in autoscaler. It would...
**Nature of request** - [x] bug report **Description** ``` go get github.com/JulienBreux/pody # k8s.io/api/core/v1 ../../../k8s.io/api/core/v1/types.go:213: undefined: v1.Initializers ``` **Environment**: ``` ⋊> ~/g/s/g/J/pody on master ◦ uname -a Linux ichbo 4.4.0-92-generic...
When running in distributed container environments (Kubernetes) resolving connection to `statsd` service may not always be successful, due to various reasons (outdated or lagging `iptables` is one of many). The...
There are times when "evict" is not sufficient to remove pods. Consider a use-case when the pod is stuck in the "Terminating" state, which could happen for various reasons, but...
Related to #1255 Currently, pod-descheduler policies are provided via config map. The policy lifecycle requires deploying a new/updated config map and restarting the pod-descheduler instance. Thus, the policy and the...
Currently, the pod-descheduler is configured with a single policy configuration, i.e., "1:1" cardinality. To support multiple policies, users must deploy multiple instances of the pod-descheduler, which poses additional operational effort...
Currently, the pod-descheduler supports the following modes of deployment/operation: - Job: one-time process - CronJob: a repeated Job invocation, i.e., repeated one-time processing. - Deployment: repeated processing via "sleep-run-sleep," where...
**What happened**: Type `kubectl` command and invoke [TAB] completion after the "-" prefix . ``` kubectl get pod -l 'app in(test)' -[TAB] ``` **What you expected to happen**: Expected to...
As a general rule, we requeue every failed request. However, there are cases when failures are non-recoverable and require an update to the resource definition itself (which in turn will...