Instances don't start stat /data/db/ps-entry.sh: no such file or directory: unknown

MongoDB Percona Operator for MongoDB
1

The instances don’t start with the error below

containerd task: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: exec: "/data/db/ps-entry.sh": stat /data/db/ps-entry.sh: no such file or directory: unknown

Although the log from the init container looks fine:

++ id -u
++ id -g
+ install -o 2 -g 2 -m 0755 -D /ps-entry.sh /opt/percona/ps-entry.sh
++ id -u
++ id -g
+ install -o 2 -g 2 -m 0755 -D /physical-restore-ps-entry.sh /opt/percona/physical-restore-ps-entry.sh
++ id -u
++ id -g
+ install -o 2 -g 2 -m 0755 -D /mongodb-healthcheck /opt/percona/mongodb-healthcheck
++ id -u
++ id -g
+ install -o 2 -g 2 -m 0755 -D /pbm-entry.sh /opt/percona/pbm-entry.sh

helm upgrade --debug --install mongodb-cluster-percona percona/psmdb-db --version 1.14.3 --namespace tenant-front  \
  --set "image.tag=6.0.2" --set "clusterServiceDNSSuffix=svc.ops-k1.domain.de" \
  --set "replsets[0].name=rs0" --set "replsets[0].size=3" \
  --set "replsets[0].volumeSpec.pvc.resources.requests.storage=2Gi" \
  --set "finalizers={delete-psmdb-pods-in-order}" \
  --set "replsets[0].resources.limits.cpu=1" \
  --set "replsets[0].resources.limits.memory=1G" \
  --set backup.enabled=false --set sharding.enabled=false
2

I am running this from a user with limited permissions, but still this user is full admin in its namespace

3

Hi @Joseph_Manel_Andres,

Welcome and thank you for joining our community. Usually this type of errors are permission related. Could you provide more information about the steps you followed and the details of the environment?

Thanks

Pep

4

Gràcies Pep!
We run multitenant environment with Capsule, and the users are able to deploy any workload on their namespaces. So I have deployed the operator with cluster admin permissions, and if I run the above helm command with cluster admin permissions, I still get the same error:

  Warning  Failed                  21s (x3 over 43s)  kubelet                  Error: failed to create containerd task: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: exec: "/data/db/ps-entry.sh": stat /data/db/ps-entry.sh: no such file or directory: unknown
  Normal   Created                 21s (x3 over 43s)  kubelet                  Created container mongod
  Normal   Pulled                  21s                kubelet                  Successfully pulled image "percona/percona-server-mongodb:6.0.2" in 845.73127ms
  Warning  BackOff                 1s (x7 over 40s)   kubelet                  Back-off restarting failed container

I have version 1.13 running fine in the same cluster, and when I tried with version 1.14 it was failing due to unable to install some CRD, which is true, we don’t allow users to install CRD but it is strange that the DB chart contained those CRD.

https://github.com/percona/percona-helm-charts/releases/download/psmdb-db-1.14.0/psmdb-db-1.14.0.tgz

5

Hi @Joseph_Manel_Andres !
In some previous helm chart versions we had crds in “-db” helm charts for some automatic testing reasons, but this was removed and now we don’t have it any more.
For psmdb-db helm chart it was removed in 1.14.2 chart version: Release psmdb-db-1.14.2 · percona/percona-helm-charts · GitHub

6

exactly, that shouldn’t be an issue anymore.