udev messages after update

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

hi all!

after the recent update to uvdev 149 i get these messages in my /var/log/messages:

katachi · n00b Joined: 12 Sep 2009 Posts: 22

I'm using a virtual hosting provider, and am experiencing something very similar. In my case, I updated from udev-141 to udev-149. I get the same "unknown key 'SYMLINK{unique}'" and "inotify_add_watch(6, (null), 10) failed" messages in /var/log/messages. I also get this on my console at boot:

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

hi!

i cannot confirm this. using udev 149 on an i686 system (virtual installation) with kernel 2.6.30-gentoo-r8 boots properly. the machine displaying the messages i posted above is an amd64 machine which is not rebooted yet (home server).

if your eth device does not work anymore try to delete this file as suggested:

ecco_the_dolphin · n00b Joined: 26 Feb 2010 Posts: 15

I have the same problem (after latest update):

ecco_the_dolphin · n00b Joined: 26 Feb 2010 Posts: 15

After I unmerged udev , "rm -rf /etc/udev/ ", and then re-emerged udev the problem is gone.

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

katachi · n00b Joined: 12 Sep 2009 Posts: 22

Decibels · Veteran Joined: 16 Aug 2002 Posts: 1630 Location: U.S.A.

I have the kernel source installed, no virtual hosting and still got the messages during boot. But everything seems to be working still.
Haven't investigated further. The system has been up since rebooted yesterday, will look into it later.
_________________
Support bacteria – they’re the only culture some people have.”

– Steven Wright

hennep · Posted: Sat Mar 13, 2010 4:42 pm Post subject:

Did you delete these files: rm /etc/udev/rules.d/70-persistent-net.rules

I got these messages after: "emerge --oneshot sys-fs/udev"

katachi · n00b Joined: 12 Sep 2009 Posts: 22

katachi · n00b Joined: 12 Sep 2009 Posts: 22

I've been testing more and have narrowed down the problem to running etc-update on /etc/init.d/udev-mount

Here are the differences:

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

katachi · n00b Joined: 12 Sep 2009 Posts: 22

More testing. Here is the problem (for me, anyway):

/etc/init.d/udev-mount

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

katachi · n00b Joined: 12 Sep 2009 Posts: 22

katachi · n00b Joined: 12 Sep 2009 Posts: 22

I have come to the conclusion that my problems are being caused by an incompatibility between recent versions of udev (greater than 141), and the Xen kernel my guest is running. While I cannot guarantee that this is indeed the problem, I have found incompatibility reports here and here. As a side note, a very helpful gentleman by the name of Andrew Lyon is providing ebuilds of more recent versions of the Xen kernel for Gentoo based on releases originally made for openSUSE. I have not tested them, but I suppose that these could possibly solve the problem for those who have the flexibility to install alternate Xen kernels.

For better or worse, I've decided to comment out the troublesome line in /etc/init.d/udev-mount, as it does not appear to negatively impact the system, but rather helps it (for now...).

That said, you guys are not running Gentoo in a virtual environment, yet are still experiencing problems. snIP3r, you indicated that you are getting your log errors on your server even before rebooting, correct? So you have not rebooted yet?

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

katachi · n00b Joined: 12 Sep 2009 Posts: 22

@snIP3r
I wish I could provide more help. I suppose that your only options are:

1) cross your fingers and reboot (hopefully you've got a backup from before the udev update). I don't know if this is possible, but maybe you could keep a downloaded version of the ebuild/source of your previously-working udev version on the home server, and fall back on that if you are unable to boot with a working net connection. I've never tried such a thing, and don't even know if it's possible.

2) move back down to the previously-working udev version before you reboot, and then mask that nothing above that version should be emerged

3) wait for someone who is more knowledgeable about this issue.

Am I correct in assuming that the log errors are not reoccurring -- that is, they appeared in the log after udev restarted, but have not appeared since then?

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

Randy Andy · Posted: Tue Mar 16, 2010 9:45 am Post subject:

Hi Folks,

i guess you have overlooked to change your kernel config accordingly, to the hints into the log.file at the beginning of this thread.

Did you all Disable CONFIG_SYSFS_DEPRECATED _and_ CONFIG_SYSFS_DEPRECATED into the kernel config?

See also here:
https://forums.gentoo.org/viewtopic-t-712186-highlight-configsysfs+deprecated.html

and work accordingly.
Then everthing should work with your packages higher than udev-149 and the belonging rules / config files.

I hope this will solve you trouble...

Andy.
_________________
If you want to see a Distro done right, compile it yourself!

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

snIP3r · l33t Joined: 21 May 2004 Posts: 853 Location: germany

hi all!

for me this issue is solved now. after re-emerging udev 149 and replacing the faulty 70-openct-rules file (from gentoo-bug tracking list #299440) i got no more error messages while restarting udev.

katachi · n00b Joined: 12 Sep 2009 Posts: 22

@snIP3r
Thank you for the follow-up!