Closed tkanng closed 5 years ago
@cofyc Can you take a look at it?
In tidb-scheduler pod, it's kube-scheduler to schedule the pod. It seems a race condition in kube-scheduler 1.12.1.
It's complex. FYI, I explained the races on the pod binding cache in pre-1.14 Kubernetes here.
Thank you so much! :+1:
duplicated with https://github.com/pingcap/tidb-operator/issues/602#issuecomment-511317477, closing this issue.
Bug Report
What version of Kubernetes are you using?
What version of TiDB Operator are you using?
What storage classes exist in the Kubernetes cluster and what are used for PD/TiKV pods?
What's the status of the TiDB cluster pods?
What did you do?
Run e2e test in Docker in Docker Kubernetes
What did you expect to see?
All clusters work.
What did you see instead?
e2e-pd-replicas-1-tikv-0
filed to start,tidb-scheduler
assigned this pod tokube-node-1
, but the local pvlocal-pv-d20c2706
is onkube-node3
. Here is the output ofkubectl describe pod e2e-pd-replicas-1-tikv-0 -n e2e-cluster2
:Here is the output of
kubectl get pv local-pv-d20c2706 -oyaml
:Here is the output of
kubectl get pv
before the e2e test:Is there something wrong with
tidb-scheduler
?:)