Proxmox 4 community updates fail postinst

[sam1994]

Hi,

I am using the PVE community repository. After running a dist-upgrade, I seem to encounter a problem which stops the packages from updating completely.

Setting up pve-manager (4.4-17) ...
Job for pvedaemon.service failed. See 'systemctl status pvedaemon.service' and 'journalctl -xn' for details.
dpkg: error processing package pve-manager (--configure):
 subprocess installed post-installation script returned error exit status 1
dpkg: dependency problems prevent configuration of proxmox-ve:
 proxmox-ve depends on pve-manager; however:
  Package pve-manager is not configured yet.

dpkg: error processing package proxmox-ve (--configure):
 dependency problems - leaving unconfigured
Errors were encountered while processing:
 pve-manager
 proxmox-ve
E: Sub-process /usr/bin/dpkg returned an error code (1)

I checked /var/lib/dpkg/info/pve-manager.postinst and it looks like this calls
/usr/bin/pveupdate. This fails on my system with the following errors:

Encode/Alias.pm did not return a true value at /usr/lib/x86_64-linux-gnu/perl/5.20/Encode.pm line 47.
BEGIN failed--compilation aborted at /usr/lib/x86_64-linux-gnu/perl/5.20/Encode.pm line 47.
Compilation failed in require at /usr/share/perl5/PVE/SafeSyslog.pm line 7.
BEGIN failed--compilation aborted at /usr/share/perl5/PVE/SafeSyslog.pm line 7.
Compilation failed in require at /usr/share/perl5/PVE/INotify.pm line 14.
BEGIN failed--compilation aborted at /usr/share/perl5/PVE/INotify.pm line 14.
Compilation failed in require at /usr/bin/pveupdate line 10.
BEGIN failed--compilation aborted at /usr/bin/pveupdate line 10.

I'd appreciate any pointers to further troubleshoot this issue.

Thanks

 

[dietmar]

Looks strange to me - maybe you modified /usr/lib/x86_64-linux-gnu/perl/5.24.1/Encode/Alias.pm ?

What is the output of

# md5sum /usr/lib/x86_64-linux-gnu/perl/5.24.1/Encode/Alias.pm

Please can you post the contents (first 100 lines) of that file?

 

[sam1994]

Hi,

Thanks for your response.

I fixed it now. I tested on another Proxmox server and had the same issue, but not on a 3rd.

It seems that if you dist-upgrade pve-manager to 4.4-15 without a reboot, then
upgrade to 4.4.17, you'll have a problem.

I had to manually grab http://download.proxmox.com/debian/...on/binary-amd64/pve-manager_4.4-15_amd64.deb; dpkg -x it over /; then tweak /var/lib/dpkg/status to reflect the old version.

After this, I was able to reboot and dist-upgrade over fine.

On both jessie boxes: Perl is 5.20.2; and md5sum is 4ed8d1759c2f3f04f28aef1cf026a11c.
First thing I did was pull perl-common and perl-modules packages and reinstall on the affected server; but didn't help.

I suspect some issues occurred in pveupdate that meant variables weren't being populated and passed
properly, so postinst bailed. If a reboot is needed between updates, I'd recommend setting /var/run/reboot-required in
postinst, and bailing in preinst if this file exists.

If I can reproduce it on other boxes, I will let you know.

Sam

 

[GadgetPig]

Hi Sam,

In case this happens again, can you try these commands if it works for you? I put descriptions in parenthesis in case anyone new to linux/proxmox are reading this:

#apt clean (cleans apt cache)
#apt update (downloads latest package lists from repositories)
#apt-get install -f (fixes broken packages)
#dpkg --configure -a (fixes interrupted dpkg installations)

Hopefully if all goes well, reboot and I usually follow up with:

#apt clean
#apt autoremove (removes packages installed by other packages and no longer needed)

BTW one thing I noticed with today's latest updates in no-subscription repository, after apt dist-upgrade, it now states:

"Seems you installed a kernel update - Please consider rebooting this node to activate the new kernel."

 

[sam1994]

Hi @GadgetPig

Thanks for the pointers. I actually maintain a Debian derivative, and we often get some interesting issues with users upgrading, so I am familiar with troubleshooting dpkg, mirroring issues and partial archives.

I tried your suggestions above before troubleshooting the postinst.

We'll see if it happens again.