[SOLVED] CT failing to start at boot - starting too early?

[sylsyl]

I installed Proxmox yesterday for the first time on new hardware. I have been using Hyper-V for about a year.

The install went well. Things appear faster than Hyper-V (but it is newer hardware).

I installed an LXC container for Plex. I have that working. It is an unpriviledged container with passthrough of the iGPU. I used the script at https://tteck.github.io/Proxmox/ to install it. I subsequently mounted my NAS hard drive using instructions I found on this site for a CIFS share. I'm really glad I read about it first or I would have ended up pulling my hair out. I have also installed a couple of test Ubuntu VMs and they appear fine.

All is working (once I increased the size of the drive in my Plex container to accommodate the cached data in the PhotoTranscoder directory) EXCEPT when I restart Proxmox, the Plex container fails to boot with the following output:

run_buffer: 571 Script exited with status 19
lxc_init: 845 Failed to run lxc.hook.pre-start for container "100"
__lxc_start: 2034 Failed to initialize container "100"
TASK ERROR: startup for container '100' failed

Interestingly, when the machine starts up I get the following in Tasks output:

"CT100 start" (which fails) followed a few seconds later by "Bulk start VMs and containers" and then immediately a "VM101 start"

I can subsequently start CT100 manually just fine.

If I tell VM101 to boot first, followed by a 10 second delay, then CT100 to boot 2nd, it starts fine.

What am I doing wrong or? It's as if CT100 is trying to start too early. What if I don't want any VMs to auto-start - how can I delay CT100?

 

[gfngfn256]

I believe you could use the Start On Boot Delay option for this purpose. See here.
You can also do it from the GUI, Node, Options, Start on boot delay & Edit.

 

[sylsyl]

Thanks - a 10 second delay seems to have made it work.

The Tasks output still suggests that the CT100 starts 4-5 seconds prior to "Bulk start VMs and Containers" but it worked this time which is the important bit. I guess it was taking a couple of seconds to mount the CIFS share which caused the trouble.