UMC "Internal server error"

Login on as Administrator, I’ve create a new user on the main UCS with “useradd -ou 0 -g 0 admin” and then later “pkill -STOP -u admin” and deluser --force.
Then all the DHCP, DNS, LDAP on the UMC give me “Internal server error: The module process died unexpectedly.”

I’m still able to access these from a backup UCS and the Active Directory on the main UCS is working fine for all the Windows server.

It only affect the Domain console on the UMC.

Please advice.

Hey,

in Linux user accounts are managed by their numeric ID, not by their name. By adding an account with the UID 0 you’ve essentially created a second root user. With that pkill you’ve sent the STOP signal to all processes with the UID 0, meaning to all processes running as root.

Have you tried sending those same processes the CONT signal? STOP only pauses the processes, it does not kill them. CONT unpauses them.

Have you tried rebooting the server?

Kind regards,
mosu

Yes, the first thing I did was to reboot the server.

It all started that as an Administrator, I couldn’t change the date, time, NTP and many of settings on the server through the GUI.
I can only do it through the terminal. With the same user, I could do all that on the backup server.

So I tried to create a new root user as admin and I couldn’t log on the console. So I forced delete it. I must have mess up somehow.

I couldn’t use any of those domain module on the UMC of the main server. It gives me Internal server error.
But the UMC on the backup server are working well. So all the changes on the LDAP, DHCP, DNS are access through the UMC of the backup server.

Everything else is working fine. I can access all the active directory from my Windows.

Hey,

interesting. An “internal server error” is usually logged by Apache. So please look at the log file /var/log/apache/error.log for further details.

Kind regards,
mosu

I’m having this authentication failure in my auth.log when accessing the LDAP on the UMC.

Apr 26 11:43:47 userver1 python2.7: nss_ldap: reconnecting to LDAP server…
Apr 26 11:43:47 userver1 python2.7: nss_ldap: reconnected to LDAP server ldap://userver1.nhksteel.local:7389 after 1 attempt
Apr 26 11:43:47 userver1 python2.7: pam_unix(univention-management-console:auth): authentication failure; logname= uid=0 euid=0 tty= ruser= rhost= user=Administrator
Apr 26 11:43:47 userver1 python2.7: pam_krb5(univention-management-console:auth): user Administrator authenticated as Administrator@NHKSTEEL.LOCAL

Please advise.

Hey,

Neither of those lines points to a problem per se:

Apr 26 11:43:47 userver1 python2.7: nss_ldap: reconnecting to LDAP server... Apr 26 11:43:47 userver1 python2.7: nss_ldap: reconnected to LDAP server ldap://userver1.nhksteel.local:7389 after 1 attempt

That nss_ldap has to reconnect sometimes is rather normal. That it was able to reconnect immediately is what counts, and that’s the case here. So this is not a problem.

Apr 26 11:43:47 userver1 python2.7: pam_unix(univention-management-console:auth): authentication failure; logname= uid=0 euid=0 tty= ruser= rhost= user=Administrator Apr 26 11:43:47 userver1 python2.7: pam_krb5(univention-management-console:auth): user Administrator authenticated as Administrator@NHKSTEEL.LOCAL

For most services the PAM configuration includes checking both the local Linux accounts via pam_unix as well as checking LDAP-based accounts by either pam_krb5 or pam_ldap. They’re tried in order. The lines shown above only state that authenticating as a local Linux user named administrator did not work (and that’s normal as administrator is an LDAP-only account), but authenticating via Kerberos did work. That’s fine, that’s expected.

Like I said above: please look at Apache’s error.log first and post it here. Thanks.

Kind regards,
mosu

Thank you for your clarifications.

Below are the /var/log/apache2/error.log

[warn] Init: Name-based SSL virtual hosts only work for clients with TLS server name indication support (RFC 4366)
[notice] Apache/2.2.22 (Univention) PHP/5.4.45-0.216.201510081527 mod_ssl/2.2.22 OpenSSL/1.0.2d mod_wsgi/3.3 Python/2.7.3 configured – resuming normal operations
[alert] (2)No such file or directory: mod_wsgi (pid=21996): Unable to change working directory to ‘/home/self-service’.
[error] [client 192.168.2.113] File does not exist: /var/www/nice ports,
[error] [client 192.168.2.66] File does not exist: /var/www/ucs-overview/js/ucs/en.json, referer: 192.168.2.3/ucs-overview/

Hey,

unfortunately that doesn’t help much. Please try to access the UMC, then go to /var/log/univention, list the files sorted by their modification date with ls -rt, and post the output of those files that have been modified since you’ve tried to access the page.

Kind regards,
mosu

These are from /var/log/univention/management-console-server.log

28.04.16 12:48:58.494 LOCALE ( WARN ) : Could not find translation file: ‘umc-core.mo’
28.04.16 12:48:58.691 LOCALE ( WARN ) : Could not find translation file: ‘diagnostic.mo’
28.04.16 12:48:58.691 LOCALE ( WARN ) : Could not find translation file: ‘udm.mo’
28.04.16 12:48:58.692 LOCALE ( WARN ) : Could not find translation file: ‘appcenter.mo’
28.04.16 12:48:58.692 LOCALE ( WARN ) : Could not find translation file: ‘join.mo’
28.04.16 12:48:58.692 LOCALE ( WARN ) : Could not find translation file: ‘ucr.mo’
28.04.16 12:48:58.692 LOCALE ( WARN ) : Could not find translation file: ‘lib.mo’
28.04.16 12:48:58.692 LOCALE ( WARN ) : Could not find translation file: ‘mrtg.mo’
28.04.16 12:48:58.693 LOCALE ( WARN ) : Could not find translation file: ‘top.mo’
28.04.16 12:48:58.693 LOCALE ( WARN ) : Could not find translation file: ‘apps.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘passwordchange.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘reboot.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘ipchange.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘updater.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘pkgdb.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘printers.mo’
28.04.16 12:48:58.694 LOCALE ( WARN ) : Could not find translation file: ‘services.mo’
28.04.16 12:48:58.695 LOCALE ( WARN ) : Could not find translation file: ‘adconnector.mo’
28.04.16 12:48:58.695 LOCALE ( WARN ) : Could not find translation file: ‘sysinfo.mo’
28.04.16 12:48:58.695 LOCALE ( WARN ) : Could not find translation file: ‘setup.mo’
28.04.16 12:48:58.695 LOCALE ( WARN ) : Could not find translation file: ‘adtakeover.mo’
28.04.16 12:48:58.695 LOCALE ( WARN ) : Could not find translation file: ‘quota.mo’
28.04.16 12:49:00.356 MAIN ( PROCESS ) : running: [’/usr/sbin/univention-management-console-module’, ‘-m’, ‘udm’, ‘-s’, ‘/var/run/univention-management-console/3819-1461822540356.socket’, ‘-d’, ‘2’, ‘-l’, ‘en_US.UTF-8’]
28.04.16 12:49:00.374 MAIN ( PROCESS ) : running: [’/usr/sbin/univention-management-console-module’, ‘-m’, ‘updater’, ‘-s’, ‘/var/run/univention-management-console/3819-1461822540373.socket’, ‘-d’, ‘2’, ‘-l’, ‘en_US.UTF-8’]
28.04.16 12:49:00.388 MAIN ( PROCESS ) : running: [’/usr/sbin/univention-management-console-module’, ‘-m’, ‘appcenter’, ‘-s’, ‘/var/run/univention-management-console/3819-1461822540388.socket’, ‘-d’, ‘2’, ‘-l’, ‘en_US.UTF-8’]
28.04.16 12:49:00.413 MAIN ( PROCESS ) : running: [’/usr/sbin/univention-management-console-module’, ‘-m’, ‘adconnector’, ‘-s’, ‘/var/run/univention-management-console/3819-1461822540412.socket’, ‘-d’, ‘2’, ‘-l’, ‘en_US.UTF-8’]
28.04.16 12:49:00.583 MAIN ( PROCESS ) : running: [’/usr/sbin/univention-management-console-module’, ‘-m’, ‘join’, ‘-s’, ‘/var/run/univention-management-console/3819-1461822540583.socket’, ‘-d’, ‘2’, ‘-l’, ‘en_US.UTF-8’]
28.04.16 12:49:02.158 MAIN ( WARN ) : Client: _recv: error on socket: [Errno 104] Connection reset by peer
28.04.16 12:49:02.158 MAIN ( WARN ) : Socket died (module=udm)
28.04.16 12:49:02.158 MAIN ( WARN ) : Module process udm died (pid: 1496, exit status: -1, signal: -1, status: -1)
28.04.16 12:49:02.158 MAIN ( WARN ) : Cleaning up requests
28.04.16 12:49:02.158 MAIN ( WARN ) : Invalidating all pending requests 146182254207614-39, 146182254028074-508
28.04.16 12:49:02.159 MAIN ( WARN ) : Remove inactivity timer
28.04.16 12:49:02.161 MAIN ( PROCESS ) : ModuleProcess: child died
28.04.16 12:49:02.161 MAIN ( WARN ) : Module process udm died (pid: 1496, exit status: -1, signal: 6, status: 6)

Hey,

thanks for the logs. Unfortunately they only show that a UDM module is crashing unexpectedly, no real hint as to the reason. Seems like deleting the user has broken something, though what is not clear at all.

At this point one would probably have to debug the UDM module. As I’m not a UDM developer I wouldn’t really know how to get started.

One last thing you could try is re-installing the UDM packages followed by a reboot:

get install --reinstall $( dpkg -l 'univention-management-console*'|awk '/^i/ { print $2 }' )

If that doesn’t help then I’m out of ideas.

Kind regards,
mosu

Mastodon