I'm running CoreOS, which uses systemd. I have several NFS mounts which work fine on my Ubuntu 14 system, but are not reliably connected at startup under systemd. I've defined them like this (where
remote is my NAS):
- name: mnt-documents.mount contents: | [Unit] After=nfs-client.target Requires=nfs-client.target [Mount] What=remote:/volume1/Documents Where=/mnt/documents Type=nfs Options=rw,vers=3 - name: mnt-documents.automount contents: | [Automount] Where=/mnt/documents
I have 4 of these, all mounting different directories from the same NFS server. Sometimes 1 fails, sometimes 2, sometimes they all work. If I try
systemctl status mnt-documents.mount I get
● mnt-documents.mount - /mnt/documents Loaded: loaded (/etc/systemd/system/mnt-documents.mount; static; vendor preset: disabled) Active: failed (Result: exit-code) since Mon 2018-02-26 09:55:24 UTC; 5h 10min ago Where: /mnt/documents What: remote:/volume1/Documents Process: 983 ExecMount=/bin/mount remote:/volume1/Documents /mnt/documents -t nfs -o rw,vers=3 (code=exited, status=32) Feb 26 09:55:23 tularemia systemd: Mounting /mnt/documents... Feb 26 09:55:24 tularemia systemd: mnt-documents.mount: Mount process exited, code=exited status=32 Feb 26 09:55:24 tularemia systemd: mnt-documents.mount: Failed with result 'exit-code'. Feb 26 09:55:24 tularemia systemd: Failed to mount /mnt/documents.
If I try to run the
mount command myself, it always succeeds.
I'm suspicious that the problem is that the mount is happening (sometimes) before DNS is active. But my question here is how to understand the failure better? I'm not sure what
status=32 is, and I'm not sure how to ask the mount-unit to pass the
-v flag to