When deploying this image as a pod in kubernetes, some CSI drivers create the mount with permissions root:PGID. All files within the mount are correctly set to PUID:PGID but since the mounting point UID is not of the PUID, the entrypoint script believes all files within do not have proper permissions. Could we add a check to see if the PGID is correct and not force the chown based solely on the PUID?
When deploying this image as a pod in kubernetes, some CSI drivers create the mount with permissions
root:PGID
. All files within the mount are correctly set toPUID:PGID
but since the mounting point UID is not of the PUID, the entrypoint script believes all files within do not have proper permissions. Could we add a check to see if the PGID is correct and not force thechown
based solely on the PUID?Thanks!