-
Notifications
You must be signed in to change notification settings - Fork 142
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
unable to power down HDD because of unreladed device still in use #1225
Comments
Could you please provide Do you see a similar error by calling the D-Bus method directly? e.g. |
Hey. Here's the dump: And no, gives the same error.
But I think I saw the problem now,... these are two identical drives, and udisks for some reason seems to identify them via ID, which are however the same for both devices:
Also in:
there is only one symlink created. There is however a Even worse,
The first call, actually powers Not sure which functionality Cheers, |
Thanks, this explains a lot. The bogus Actually, your setup may be classified as a multipath scenario when multipath-aware tools are considered (do not install any as you may corrupt your data). While there's some basic support of multipath in UDisks, it's mostly harmless. Anyway, may I ask you for As a tentative proposed workaround, UDisks may want to ignore multipathing on devices connected over usb. |
Well, but for udev, and most of the
Well these are two distinct devices, and I'd consider any software severely broken, that relies on e.g. UUID, label, serial numbers or the likes and not the true device major/minor numbers or something that's guaranteed to be unique.
Will take a while, because I'm not at home.
And as I've said before... depending on what udisks offers in terms of exported functionality (like wiping filesystems, creating new ones, etc. pp.), using anything that is not guaranteed - from the kernel side - to be unique, may easily lead to catastrophic data loss. I also doubt that just applying a workaround for USB will be a proper solution. While these USB-SATA bridges are often cheap devices and they're therefore likely to do such stuff... there's no guaranteed that something like this may not also happen on SATA controller connected by other means (or just for any other block device, connected somehow). |
Hi, I personally also ran into this issue and am spotting something weird. My
This value
I hope this helps in finding the potential issue, but to me it looks like there's some problems with not working well when two of the same drives are connected.. This is on version 2.9.4 If that helps |
@Galdanwing can you provide |
Sure, I'm now on v2.10.1 and still facing it:
udevadm /dev/sdd:
and of sdh:
|
I have now also tried to do it via any of the other /dev/disk/ links, but they all fail. Hope that helps! |
Hey.
This must have started with a recent upgrade of
udisks
or maybe one of its libs likelibblockdev
(unfortunately I don't know exactly which).What I do is:
For example, if
sda
is the one where everything is stopped/unmounted:The latter makes sense, as the LUKS on
sdb1
is actually still running, however, even when trying to power offsda
, which has no business withsdb1
, it complains aboutsdb1
still being unlocked.Any ideas?
Thanks,
Chris.
The text was updated successfully, but these errors were encountered: