To access some scientific software, I have been running a script on a remote share accessed through autofs and automount. Everything was running fine for months when this script started to hang and it looks like automount is the problem.
When trying to run the script I try to "execute" a symbolic link pointing to the real script at the remote location that automount should mount but this is what happens:
Starting autofs :
service autofs start
dmesg output :
SELinux: initialized (dev autofs, type autofs), uses genfs_contexts
automount: lookup_read_master: lookup(nisplus): couldn't locate nis+ table auto.master
Then I try to execute the script, command seems to hang, prompt returns only when I hit <CTRL+C>.
dmesg output at this point :
automount: segfault at 28 ip 00007f285e149862 sp 00007f28619e1960 error 4 in lookup_hosts.so[7f285e140000+1c000]
kernel: automount: segfault at 28 ip 00007fab18299862 sp 00007fab1b12f960 error 4 in lookup_hosts.so[7fab18290000+1c000]
abrtd: Directory 'ccpp-2012-08-08-22:13:17-16806' creation detected
abrt: Saved core dump of pid 16806 (/usr/sbin/automount) to /var/spool/abrt/ccpp-2012-08-08-22:13:17-16806 (24166400 bytes)
9-10 seconds later some more appears:
abrtd: Sending an email...
abrtd: Email was sent to: root@localhost
abrtd: Duplicate: UUID
abrtd: DUP_OF_DIR: /var/spool/abrt/ccpp-2012-07-21-16:35:37-2125
abrtd: Problem directory is a duplicate of /var/spool/abrt/ccpp-2012-07-21-16:35:37-2125
Deleting problem directory ccpp-2012-08-08-22:13:17-16806 (dup of ccpp-2012-07-21-16:35:37-2125)
Any help will be greatly appreciated!