mount.cifs + krb5 looking for ticket at incorrect location
edward2a opened this issue · comments
What happened:
Mount failed and container crashing.
May 11 10:19:42 worker-02 cifs.upcall: get_existing_cc: default ccache is FILE:/tmp/krb5cc_0
What you expected to happen:
Mount successful and container started.
How to reproduce it:
Install using manual installation (described below).
Anything else we need to know?:
The smb container is failing to mount with the error above.
The host has the appropriate kerberos configuration as per the documentation, pointing the krb5cc to
/var/lib/kubelet/kerberos/krb5cc_%{uid}.
Mounting from the host works appropriately with the exact same command as issued by the smb container.
Upon adding a krb5.conf inside the container with the appropriate configuration for the krb5cc path the mount executes correctly.
Environment:
Kubernetes 1.25 cluster (1 master, 3 workers)
Windows Server 2019 AD
- CSI Driver version: master (see below)
- Kubernetes version (use
kubectl version): 1.25 - OS (e.g. from /etc/os-release): debian 11
- Kernel (e.g.
uname -a):Linux 5.10.0-22-amd64 #1 SMP Debian 5.10.178-3 (2023-04-22) x86_64 GNU/Linux - Install tools:
deploy/install-driver.sh master local - Others:
image: gcr.io/k8s-staging-sig-storage/smbplugin canary 87ca29f8edd3e 77.4MB
Build Date: "2023-05-08T23:15:20Z"
Compiler: gc
Driver Name: smb.csi.k8s.io
Driver Version: v1.10.0
Git Commit: ""
Go Version: go1.19
Platform: linux/amd64
It does indeed, thanks for the reminder!