Weird problem with middleware

[NickF]

Hi all,

I've been having a weird bug with the middleware. I can report a jira ticket, but I'm not sure exactly what logs would be relevant or information which may be useful.

Every few days in the webui my virtual machines appear as though they all got turned off. But they didn't. If I restart the middleward daemon from the CLI it fixes the problem and properly shows what VMs are up or down in the UI. However, VNC port mapping gets broken and I need to restart my VMs in order for it to work again.

Any ideas so I can more cohesively report the problem are welcome.

 

[morganL]

Which version of software.?
Did it work on previous versions?

 

[zizzithefox]

Which version of software.?

Did it work on previous versions?

Same problem here.
Upgraded to 13-U1, now 13-U2, and I am about to upgrade to 13-U3 in the weekend and see how it goes.
Being lazy, I thought someone was going to report it and didn't do it myself. Besides, it is an old machine (Xeon E3-1225, the other one is a slightly newer E3-1230v3, but I am still on 12 with that, and it's not running 24/7).
I rarely have the need to restart the machines, as they are working despite the fact they are reported as powered off, so it's not affecting me significantly, but it can be annoying.

I am 90% positive it is something related to 13, but not entirely sure.

How can I provide some information about what happened (maybe middlewared.log)? I had a look and I have a lot of python errors when it says that it was unable to retrieve the state of the machines...

[Edit]
Nothing changed with 13-U3

 

[BitsAndBytes]

Same here - running TrueNAS-13.0-U3.1 (Core not Scale).
Restarting middleware gets things working again.
Small extra point - VNC to vm also is also unavailable until middleware restarted.

Is there a Jira yet?