View previous topic :: View next topic |
Author |
Message |
friesia Apprentice
Joined: 23 Mar 2007 Posts: 202
|
Posted: Mon Dec 03, 2012 5:43 pm Post subject: Problems with LVM utilities on systemd |
|
|
Here I'm trying to create a PV, VG and then a LV.
Let me say that this only happens when I boot into systemd. With OpenRC everything's OK. I'm on ~amd64 and use all the latest software.
Code: |
melforce ~ # pvcreate /dev/sda4
Physical volume "/dev/sda4" successfully created
melforce ~ # vgcreate derk /dev/sda4
Volume group "derk" successfully created
melforce ~ # vgdisplay
--- Volume group ---
VG Name siga
<I skipped some lines here>
--- Volume group ---
VG Name derk
System ID
Format lvm2
Metadata Areas 1
Metadata Sequence No 1
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 0
Open LV 0
Max PV 0
Cur PV 1
Act PV 1
VG Size 44,43 GiB
PE Size 4,00 MiB
Total PE 11373
Alloc PE / Size 0 / 0
Free PE / Size 11373 / 44,43 GiB
VG UUID k2D7Wp-VJT8-25yG-qhnF-qfDX-xFZx-c6E1yf
melforce ~ # lvcreate -L5G -nmylv derk
/dev/derk/mylv: not found: device not cleared
Aborting. Failed to wipe start of new LV.
semid 9797633: semop failed for cookie 0xd4d0a9b: incorrect semaphore state
Failed to set a proper state for notification semaphore identified by cookie value 223152795 (0xd4d0a9b) to initialize waiting for incoming notifications.
|
Thus, LV is not created.
If I use --zero n it's created alright:
Code: |
melforce ~ # lvcreate -Z n -L5G -nmylv derk
WARNING: "mylv" not zeroed
Logical volume "mylv" created
|
But still I get semaphore warnings on delete:
Code: |
melforce ~ # lvremove /dev/derk/mylv
Do you really want to remove active logical volume mylv? [y/n]: y
Logical volume "mylv" successfully removed
semid 10780673: semop failed for cookie 0xd4db599: incorrect semaphore state
Failed to set a proper state for notification semaphore identified by cookie value 223196569 (0xd4db599) to initialize waiting for incoming notifications.
|
I have no idea what it can be related to. |
|
Back to top |
|
|
Mardok45 n00b
Joined: 21 Jun 2008 Posts: 69 Location: Right behind you
|
Posted: Mon May 20, 2013 9:00 pm Post subject: |
|
|
Dear people from the future who happen to be searching for the same problem:
I was experiencing the same problem and eventually found this:
http://www.redhat.com/archives/linux-lvm/2012-August/msg00025.html
After changing the following entry in the kernel .config from:
Code: | CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" |
to
Code: | CONFIG_UEVENT_HELPER_PATH="" |
the problem seems to be fixed. |
|
Back to top |
|
|
Laetos n00b
Joined: 18 Jan 2003 Posts: 27 Location: utopia
|
Posted: Fri Jul 26, 2013 9:36 am Post subject: |
|
|
Hi Mardok45 the Future is here, just stopping by to say that this did help me fix it, thx! |
|
Back to top |
|
|
glaukommatos n00b
Joined: 05 Jan 2014 Posts: 3
|
Posted: Sun Jan 19, 2014 7:09 am Post subject: |
|
|
Thanks, this worked for me too! |
|
Back to top |
|
|
ub1quit33 n00b
Joined: 24 May 2013 Posts: 16
|
Posted: Sun Feb 09, 2014 10:56 pm Post subject: |
|
|
Much thanks for sharing your solution. This fixed my issues with LVM under systemd. |
|
Back to top |
|
|
TomWij Retired Dev
Joined: 04 Jul 2012 Posts: 1553
|
Posted: Mon Feb 10, 2014 8:38 am Post subject: |
|
|
I'm not sure which package this is about; but to make sure future users don't experience this again, could you file a bug indicating that the package in question needs to check for this? (Or let me know the package?) |
|
Back to top |
|
|
psofa Guru
Joined: 28 Feb 2004 Posts: 485
|
Posted: Tue Oct 07, 2014 1:44 am Post subject: |
|
|
Thank you!
can also be fixed at runtime with:
echo 0 > /sys/kernel/uevent_helper
I guess "sys-fs/lvm2" should be checking for this _________________ psofa |
|
Back to top |
|
|
steveL Watchman
Joined: 13 Sep 2006 Posts: 5153 Location: The Peanut Gallery
|
Posted: Tue Oct 07, 2014 2:09 am Post subject: |
|
|
That path has been deprecated for a long time, since running a helper process can be racy. netlink is a much better interface, though it does require some specific coding.
afaik, only mdev still uses that path, and ofc the capability is there for someone building their own kernel, to hook into whatever they want.
I could have sworn I read about that in the kernel config documentation, as well as other places wrt to udev changes (two or three years ago.) |
|
Back to top |
|
|
AchilleTalon Guru
Joined: 11 Apr 2004 Posts: 368 Location: Montreal, Quebec, Canada
|
Posted: Tue Nov 04, 2014 12:23 am Post subject: [solved] |
|
|
I am adding here my question since it is very similar and I am not sure if it is a different problem. I did recompiled the kernel removing the helper and it didn't fix my problem.
Here is the message I am getting:
Code: | Volume "donneesvg/db2data" is not active locally.
Aborting. Failed to wipe start of new LV.
|
I must say at boot time, the boot process is throwing me into a maintenance shell where I have to vgchange -a y donneesvg to activate the volume group, then systemctl default to continue the boot process. I do have a discussion thread on this problem, nobody has came so far with anything useful to try to fix this problem. So, I don't know if my two problems are related and simply two different symptoms of a single problem.
Ignore the italic part above, the problem is unrelated and I fixed this problem which has something to do with kernel options for dracut.
What does it mean "not active locally"? I tried to find something about it on the web and found nothing.
NOTE: Solved, my problem was the new LV was not included in the volume_list. _________________ Achille Talon Hop! |
|
Back to top |
|
|
|