I couldn't think of the best way to write this. Apologies in advance.
The setup:
PMS on Centos 7 minimal within ESXI (6.5.0U1)(2 vcpu, 8gb ram at the given time)
Using NFS to connect to multiple synology 1817+ with future expansion plans (may or may not be synology).
The why:
I wanted plex's uid of 996 to match on the multiple NAS. I could edit this by way of /etc/passwd (no usermod avail) on each of the individual NAS, but didn't feel content with this method. The uid of plex on each NAS wasn't guaranteed to be the same with each other so replacing the uid on the plex vm didn't seem feasible. Not to mention, when I did attempt this: plex broke. oops!
The bandage:
First of all: thank you to everyone on this forum.
I didn't really want a plex user on each of the NAS and already had a RO user (all using the same uid) that I was using. I stopped the plex service, added the following to an override.conf:
$ cat /etc/systemd/system/plexmediaserver.service.d/override.conf
#
[Service]
User=nayeon
Group=nayeon.
#
$
chown'd /usr/lib/plexmediaserver to nayeon:nayeon, used daemon-reload, and restarted the plex server.
Using one NAS as an example, its fstab entry is:
192.168.1.7:/volume1/korean /mnt/nfs/korean nfs defaults,sec=sys,ro,async,bg 0 0
Within synology's ui, nayeon (uid is 1027 on all machines) is given RO permission to the shared folder. Group permissions didn't seem like a good route.
Where I'm at now:
When using the plex user, it wouldn't have access to the folder. (only root did initially) By using nayeon, the folder contents are now accessible but I feel it may not be stable. When setting up a library, plex front-end slowed to a crawl and occasionally gave "There was an unexpected error loading the dashboard" errors. To be specific, this was while fstab was using defaults 0 0. I haven't had enough time to see if the added entries made any changes. The reality is: I couldn't find a lot of current information on changing away from the default plex user and want to make sure everything necessary was done (or if there is a better way to do it from plex vm side only)
NFS rule:
hostname or ip: 192.168.1.51
privilege: read only
squash: no mapping
security: sys
enable async checked
allow connections from port 1024+ unchecked
allow users to access mounted subfolders checked
