dhcpcd: Dependent NFS mounts fail/never happen

[Athanasius]

On Tue, Nov 05, 2024 at 02:30:04PM +0000, Athanasius wrote:
>   Yes, utilising `x-systemd.requires=network-online.target` in my NFS
> mount /etc/fstab lines does indeed appear to have resolved this.  I can
> even see the correct ordering (systemd NFS mount units only firing once
> "Reached target network-online.target - Network is Online." is reported)
> in journalctl output.

  Whilst *this* bug is definitely against the wrong package...

  ... this morning my not-$HOME NFS mount failed again.  It looks like
some race between whatever systemd is doing to detect/declare
`network-online.target` success and when dhcpcd has actually fully
gotten the network configured.  The logging *looks* like it should have
worked, but it didn't:

Nov 06 08:26:09 emilia ifup[1401]: lan0: offered 192.168.1.122 from 192.168.1.1
Nov 06 08:26:09 emilia dhcpcd[1401]: lan0: offered 192.168.1.122 from 192.168.1.1
Nov 06 08:26:09 emilia ifup[1401]: lan0: ignoring offer of 192.168.1.122 from 192.168.1.1
Nov 06 08:26:09 emilia dhcpcd[1401]: lan0: ignoring offer of 192.168.1.122 from 192.168.1.1
Nov 06 08:26:09 emilia ifup[1401]: lan0: probing address 192.168.1.122/24
Nov 06 08:26:09 emilia dhcpcd[1401]: lan0: probing address 192.168.1.122/24
...
Nov 06 08:26:13 emilia systemd[1]: Finished networking.service - Raise network interfaces.
Nov 06 08:26:13 emilia systemd[1]: Reached target network.target - Network.
Nov 06 08:26:13 emilia systemd[1]: Reached target network-online.target - Network is Online.
Nov 06 08:26:13 emilia systemd[1]: Mounting home-users.mount - /home/users...
Nov 06 08:26:13 emilia systemd[1]: Mounting mnt-tuesday-data0.mount - /mnt/tuesday/data0...
...
Nov 06 08:26:13 emilia systemd[1]: mnt-tuesday-data0.mount: Mount process exited, code=exited, status=32/n/a
Nov 06 08:26:13 emilia systemd[1]: mnt-tuesday-data0.mount: Failed with result 'exit-code'.
Nov 06 08:26:13 emilia systemd[1]: Failed to mount mnt-tuesday-data0.mount - /mnt/tuesday/data0.
Nov 06 08:26:13 emilia mount[5063]: mount.nfs: Network is unreachable
Nov 06 08:26:13 emilia systemd[1]: home-users.mount: Mount process exited, code=exited, status=32/n/a
Nov 06 08:26:13 emilia systemd[1]: home-users.mount: Failed with result 'exit-code'.
Nov 06 08:26:13 emilia systemd[1]: Failed to mount home-users.mount - /home/users.
...
Nov 06 08:26:13 emilia dhcpcd[1401]: lan0: leased 192.168.1.122 for 3600 seconds
Nov 06 08:26:13 emilia avahi-daemon[1433]: Joining mDNS multicast group on interface lan0.IPv4 with address 192.168.1.122.
Nov 06 08:26:13 emilia avahi-daemon[1433]: New relevant interface lan0.IPv4 for mDNS.
Nov 06 08:26:13 emilia avahi-daemon[1433]: Registering new address record for 192.168.1.122 on lan0.IPv4.
Nov 06 08:26:13 emilia dhcpcd[1401]: lan0: adding route to 192.168.1.0/24
Nov 06 08:26:13 emilia dhcpcd[1401]: lan0: adding default route via 192.168.1.1

  Now, this is *possibly* because I also have IPv6 networking and that
*is* fully configured before "Reached target network-online.target -
Network is Online.".  I'll see about experimenting with adjusting my NFS
mounts to either use IPv6 instead, or just the hostname.
  But, it still seems valid to me that an NFS mount *explicitly
configured to use an IPv4 address* should not be attempted by systemd
until *IPv4 networking is fully up*.

  This is definitely more a systemd bug than anything else.
-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
                  GPG/PGP Key: https://miggy.org/gpg-key
	   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME