I raised another issue about signal-event interface-update hanging and then causing my system to fail to reconfigure, and then basically become unavailable.
After digging through, I found out why: My Nextcloud has a data directory that is on a network drive.
Here we can see the sequence of interface-update
- S03nethserver-alerts-reset-wan
- S04network-stop
- S20interface-rename
- S21adjust-fixnetwork-flag
- S25interface-config-reset
- S30interface-config-write
- S50nethserver-nextcloud-occ-conf
- S60nethserver-dnsmasq-adjustdb
- S61nethserver-dnsmasq-signalsave
- S70network-start
- S80static-routes-save
- S90firewall-adjust
- S95trusted-networks-modify
- S99nethserver-docker-interface-update-restart-docker
The Nextcloud configuration happens while the network is down, and even the simple occ commands for adjusting Nextcloud configuration will fail while the data directory is unavailable.
I found that changing S50nethserver-nextcloud-occ-conf to S91nethserver-nextcloud-occ-conf makes everything work just fine.
Is it possible to get this updated in the default install? Also, could the script be improved so that if it crashes it won’t stop the rest of the configuration steps from proceeding?