I have a Gluster Cluster which is mounted onto a Linux machine that serves it over NFS.
I have a Git Repo stored on this cluster, but occasionally the system has an issue with locking on the files.
i.e. when I go to commit, I get <path>/<branch>.lock cannot be renamed to <path>/<branch>: Try again?
Sometimes restarting the NFS service on the server fixes it. Sometimes restarting the Gluster service fixes it. And sometimes I have to delete the original file after making a backup copy and moving he backup into its original location which fixes it.
Is there a better way to handle this?