Reboot on TrueNAS SCALE 21.06-BETA seams to take forever

[oumpa31]

Has anyone else noticed that TrueNAS SCALE 21.06-BETA takes like 20+ minutes to restart. It seems to slow down to a crawl when its importing the pools.

 

[mattzera]

I have not noticed this... Hopefully you can find a fix. My system boots in ~3 or 4 minutes

 

[morganL]

Has anyone else noticed that TrueNAS SCALE 21.06-BETA takes like 20+ minutes to restart. It seems to slow down to a crawl when its importing the pools.

Can you upgrade to 21.06 BETA and document you system setup?

 

[oumpa31]

My system is a bit big, I am on 21.06 BETA
AMD Ryzen Threadripper 2970WX 24-Core
128 Gigs ram
(2) 256 Gig nvme m.2 boot drive mirrored boot
Quadro P4000
Intell X540-T2 aggerated in lacp
LSI 9201-16e attached to (2) 24bay disk shelves
2.5" disk shelf
(3) 2tb ssd raidz App and VM pool
(9) 2tb ssd raid3 Video editing pool
(7) 5tb HHD camera / backup pool
3.5" disk shelf
(12) 4tb drives raidz3
(12) 14tb drives raidz3
combined to make one giant main pool

 

[Muddro]

I have had a similar issue though may not be the same issue. On reboot for me, when looking at top in shell, asyncio_loop seems to hang and eat CPU. If you notice after 10 minutes after rebooting that it's the top process it may be the culprit. I have ended up killing the process after reboot which allows it to start over. It gets back to 100% Cpu. However, after about 4 minutes the CPU usage drops off and everything is working as it should. Fyi, for a majority of those 4 minutes the webui is inaccessible as well.

 

[oumpa31]

I have a screen attached to my server and this is what i get when it is booting up you can see at this time it had been 14 mins and some change. The web ui doesn't work at this time at all.

 

[morganL]

If you think its a bug, please report it with your hardware configuration.

 

[oumpa31]

I think I may have to do that. I have to do 1 more restart now that all my pools no longer have a dive with errors to see if that was the problem or if this happens all the time.

 

[fsch]

I think I may have to do that. I have to do 1 more restart now that all my pools no longer have a dive with errors to see if that was the problem or if this happens all the time.

Have you already reported it? It happens for me too, boot takes around 6minutes for me, spending most time at the same message.

My System:
- Intel Core i5-9400F
- 16GB DDR4 RAM
- 120GB M.2 SATA SSD as boot-pool
- 2x 4TB Seagate Ironwolf as main data pool
- 2x 2TB SATA HDD as secondary pool
- GTX 750 Ti as GPU for VMs
- Radeon HD 5450 as GPU for Server
- TrueNAS-SCALE-21.06-BETA.1

If you have reported it, could you link me to the jira ticket?

 

[ornias]

When having loads of containers this might happen, is that the case here as well?

 

[stavros-k]

When having loads of containers this might happen, is that the case here as well?

I haven't ever used ZFS in the past so... Genuine question:
Why would the number of containers affect the pool import?

I get that it will take time to deploy the containers, but slow import of pools seems weird to me.
But I would also accept slow import if someone had a large number of pools.

Both my vm and my barebones machine, boot (after an update) in ~5min, (I can get to login screen), and then another ~5min to fully deploy 4-5apps

 

[fsch]

I'm not using any "built-in" apps, I've got a few containers running via docker-compose. I don't think that is the cause of the problem, as far as I remember it booted faster with the previous alpha version(s) of SCALE. Might try comparing that later.

 

[ornias]

@StavrosMadK Pool import shouldn't really be affected, however: I have noticed slow shutdown with a lot of containers.
Although that might mostly be networking stack related though...