Lots of 'failed to remove ioctl... luks' lines spamming on shutdown. Is it dangerous?

[tonsimple]

Qubes OS version (e.g., R3.2):

3.2

Affected TemplateVMs (e.g., fedora-23, if applicable):

Dom0
(I think...)

---

Expected behavior:

Clean shutdown

Actual behavior:

late in shutdown there is a pause and many (maybe hundreds!) of messages like :

device-mapper: remove ioctl on [something I can't read because scrolls to fast but has word luks in it] failed [something] device or resource busy

After a while system still shutdowns

These messages sound scary.
Should I be worried ?

Steps to reproduce the behavior:

1. shut down all VMs (so we're sure it's not a VM hanging)
2. run system for like five hours
3. trigger system shutdown from the GUI

General notes:

Doesn't always happen.
Sometimes.
Usually if uptime > 5 hours.
Pointers as to how to find out what it is and how dangerous it is are very appreciated.

---

Related issues:

[andrewdavidwong]

I see these too, and I'd also like to know. Any idea, @marmarek?

[andrewdavidwong]

I see these too, and I'd also like to know. Any idea, @marmarek?

[pdinoto]

I see these too, and seems to be related to shutdown when there was at least one machine with an attached block device.

In my case, dom0 root disk is on a SSD, and I have a 2nd LUKS volume at /var/lib/qubes-add which is a SHDD disk and added as a xen pool.

Will try to film the issue, as it happens quite frequently for me.

[pdinoto]

I see these too, and seems to be related to shutdown when there was at least one machine with an attached block device.

In my case, dom0 root disk is on a SSD, and I have a 2nd LUKS volume at /var/lib/qubes-add which is a SHDD disk and added as a xen pool.

Will try to film the issue, as it happens quite frequently for me.

[marmarek]

It's probably related to long system shutdown ( #1581 ), I guess VMs are not stopped and are still connected to those devices. I still haven't figured out why VMs are not stopped at this point...

[marmarek]

It's probably related to long system shutdown ( #1581 ), I guess VMs are not stopped and are still connected to those devices. I still haven't figured out why VMs are not stopped at this point...

[andrewdavidwong]

@marmarek: I don't think that's what it is (at least not in my case), because I make sure to shut down all VMs before issuing a shutdown/reboot command in dom0.

[andrewdavidwong]

@marmarek: I don't think that's what it is (at least not in my case), because I make sure to shut down all VMs before issuing a shutdown/reboot command in dom0.