[opensuse-factory] LVM logical volume not activated automatically
The same setup as mine and the same problem after the update. however i think the solution is just a coincidence since I had both packages installed already. WHat I have noticed howeever is that I get two jobs running one for udev (udev-settle) with a limit of 180s and the one for lvm activation of 90s. now i speculate that whatever settle is doing has effect on lvm activation and is not finished in the 90s. I reduced the time in settle to 70s and after approx 2min i have the login screen up without the emergency systemd. systemd-analyze blame 1min 50.210s lvm2-activation-net.service 1min 10.197s systemd-udev-settle.service 8.664s systemd-cryptsetup@cr_nvme \x2d2002538435b1b37d4\x2dpart2.service my understanding is this can be cause by many things so once i get more time I may dig more. Alin -- Without Questions there are no Answers! _______________________________________ Dr. Alin Marin ELENA http://alin.elena.space/ _______________________________________
On Tue, 2017-01-03 at 17:35 +0000, Alin M Elena wrote:
The same setup as mine and the same problem after the update. however i think the solution is just a coincidence since I had both packages installed already.
WHat I have noticed howeever is that I get two jobs running one for udev (udev-settle) with a limit of 180s and the one for lvm activation of 90s.
now i speculate that whatever settle is doing has effect on lvm activation and is not finished in the 90s. I reduced the time in settle to 70s and after approx 2min i have the login screen up without the emergency systemd.
systemd-analyze blame 1min 50.210s lvm2-activation-net.service 1min 10.197s systemd-udev-settle.service 8.664s systemd-cryptsetup@cr_nvme \x2d2002538435b1b37d4\x2dpart2.service
my understanding is this can be cause by many things so once i get more time I may dig more.
lvm2-activation-net.service runs "vgchange -aay" after activation of "remote" (iscsi and fcoe) block devices. If this service hangs for so long, you may be facing some problem with one ore more devices not responding (not necessarily iscsi or fcoe devices). The fact that udev settle takes so long probably has a similar reason. You may want to paste "journalctl -ab" output somewhere. Martin -- Dr. Martin Wilck <mwilck@suse.com>, Tel. +49 (0)911 74053 2107 SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton HRB 21284 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Thanks Martin, here it is. http://sprunge.us/gfHW I suspect something goes wrong at shutdown. since i see running jobs runnig related to the lvm attached bellow. http://sprunge.us/efbg Alin On Thursday, 5 January 2017 20:43:02 GMT Martin Wilck wrote:
On Tue, 2017-01-03 at 17:35 +0000, Alin M Elena wrote:
The same setup as mine and the same problem after the update. however i think the solution is just a coincidence since I had both packages installed already.
WHat I have noticed howeever is that I get two jobs running one for udev (udev-settle) with a limit of 180s and the one for lvm activation of 90s.
now i speculate that whatever settle is doing has effect on lvm activation and is not finished in the 90s. I reduced the time in settle to 70s and after approx 2min i have the login screen up without the emergency systemd.
systemd-analyze blame 1min 50.210s lvm2-activation-net.service 1min 10.197s systemd-udev-settle.service 8.664s systemd-cryptsetup@cr_nvme \x2d2002538435b1b37d4\x2dpart2.service
my understanding is this can be cause by many things so once i get more time I may dig more.
lvm2-activation-net.service runs "vgchange -aay" after activation of "remote" (iscsi and fcoe) block devices. If this service hangs for so long, you may be facing some problem with one ore more devices not responding (not necessarily iscsi or fcoe devices). The fact that udev settle takes so long probably has a similar reason. You may want to paste "journalctl -ab" output somewhere.
Martin
-- Without Questions there are no Answers! _______________________________________ Dr. Alin Marin ELENA http://alin.elena.space/ _______________________________________
On Thu, 2017-01-05 at 22:36 +0000, Alin M Elena wrote:
Thanks Martin, here it is. http://sprunge.us/gfHW
Is this a regression with the 4.10-rc2 kernel? As a workaround, you could try to blacklist the aesni_intel module which is apparently causing trouble: echo blacklist aesni_intel >>/etc/modrobe.d/aesni.conf Background: I can see a BUG() in your logs related to loading the aesni_intel module. It's probably caused by recent changes in the crypto layer of the kernel. The boot delay is related to the following messages: Jan 05 18:32:01 circassia systemd-udevd[805]: seq 2404 '/devices/system/cpu/cpu0' is taking a long time Jan 05 18:32:01 circassia systemd-udevd[805]: seq 2407 '/devices/system/cpu/cpu3' is taking a long time ... Jan 05 18:34:01 circassia.dl.ac.uk systemd-udevd[805]: seq 2404 '/devices/system/cpu/cpu0' killed Jan 05 18:34:01 circassia.dl.ac.uk systemd-udevd[805]: seq 2407 '/devices/system/cpu/cpu3' killed ... Jan 05 18:34:01 circassia.dl.ac.uk systemd-udevd[805]: worker [994] terminated by signal 9 (Killed) Jan 05 18:34:01 circassia.dl.ac.uk systemd-udevd[805]: worker [994] failed while handling '/devices/system/cpu/cpu0' Jan 05 18:34:01 circassia.dl.ac.uk systemd-udevd[805]: worker [995] terminated by signal 9 (Killed) Jan 05 18:34:01 circassia.dl.ac.uk systemd-udevd[805]: worker [995] failed while handling '/devices/system/cpu/cpu1' Udev worker hanging on a CPU "add" or "change" event - I've never seen this before. Udev has hardly anything to do for cpu devices... but it's probably a side-effect of the BUG above.
I suspect something goes wrong at shutdown. since i see running jobs runnig related to the lvm attached bellow. http://sprunge.us/efbg
Not sure what you mean. I can see problems with your encrypted disk, which is again probably related to the BUG(). Martin -- Dr. Martin Wilck <mwilck@suse.com>, Tel. +49 (0)911 74053 2107 SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton HRB 21284 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Is this a regression with the 4.10-rc2 kernel? it may be.
As a workaround, you could try to blacklist the aesni_intel module which is apparently causing trouble:
echo blacklist aesni_intel >>/etc/modprobe.d/aesni.conf this indeed did the trick,
here is the new log http://sprunge.us/aMVT you were right on side effects related to cpu now they seem to be gone. Thanks! Alin -- Without Questions there are no Answers! _______________________________________ Dr. Alin Marin ELENA http://alin.elena.space/ _______________________________________
participants (2)
-
Alin M Elena -
Martin Wilck