Bugzilla – Bug 426396
nscd not running
Last modified: 2008-11-19 19:04:02 UTC
nscd is not running. When I check /var/log/messages, I see: Sep 14 19:54:04 nwr2 nscd: 7082 invalid persistent database file "/var/run/nscd/passwd": verification failed Sep 14 19:54:04 nwr2 nscd: 7082 invalid persistent database file "/var/run/nscd/services": Permission denied Sep 14 19:54:04 nwr2 nscd: 7082 cannot write to database file /var/run/nscd/services: Permission denied Sep 14 19:59:55 nwr2 nscd: 7171 cannot write to database file /var/run/nscd/services: Permission denied Sep 15 08:20:28 nwr2 nscd: 9929 cannot write to database file /var/run/nscd/services: Permission denied Sep 15 08:23:42 nwr2 nscd: 9998 cannot write to database file /var/run/nscd/services: Permission denied Those are all since the last boot. Further back (from an earlier boot), I also see: Sep 8 08:30:15 nwr2 kernel: nscd[2994]: segfault at ff8582f4 ip b804b7dc sp afc82fbc error 4 in nscd[b803b000+1c000] This may be the same bug as #371582, except this is for opensuse 11.0 and the previously reported bug was for opensuse 10.3. I thought it better to generate a new report due to different version. The "/etc/nscd.conf" file is as originally installed. The two log messages from Sep 15 (listed above) suggest that something is attempting to restart nscd. However, it still is not running. Incidently, the file "/var/run/nscd/services" does not exist.
Can you run ulimit -c unlimited, run nscd manually with -d and report what message it crashed on and attach the core?
Ran with "ulimit -c unlimited", starting manually. "nscd" ran for a while, then stopped running. There was no core. (I looked in "/", which "lsof" had shown as the current directory for the running process.) An additional comment is in order. The message Sep 14 19:54:04 nwr2 nscd: 7082 cannot write to database file and the message Sep 14 19:59:55 nwr2 nscd: 7171 cannot write to database file appear to be related to starting and stopping a cisco vpn client. Presumably the vpn client is attempting to restart nscd at both points. Note that while testing with "ulimit -c unlimited", I avoided starting the vpn client. That nscd stops running appears unrelated to the vpn. I am seeing nscd fail on two different computers with very different hardware.
information was provided.
Are you sure you started nscd with the -d flag? What was the last message nscd emitted before it stopped running? (It should be quite chatty with the -d flag.)
Oops! I missed that "-d". Sorry about that. Here are the last few lines of output before it stopped running: 7111: Reloading "200" in group cache! 7111: Reloading "100" in group cache! 7111: remove GETPWBYUID entry "0" 7111: remove GETPWBYNAME entry "root" nscd: mem.c:392: gc: Assertion `off_alloc == off_allocend' failed. Abort There was no coredump.
Hmm, the bug is set to be in 11.1 but in your initial comment you mention 11.0. What version are you using?
I am using 11.0. Apologies if I misposted
Ok, then we track this bug already. *** This bug has been marked as a duplicate of bug 387202 ***