I recently updated a Red Hat 7 host. After rebooting, the nfs mounts on
it (nfsver3,sec=krb5) failed to mount, due to the gss daemon segfaulting
when the mount attempt happened:
[ 7.816487] FS-Cache: Loaded
[ 7.887575] FS-Cache: Netfs 'nfs' registered for caching
[ 7.931164] rpc.gssd[498]: segfault at 5544452e ip 00007fbc9d704ee6
sp 00007ffc37291678 error 4 in libc-2.17.so[7fbc9d5ca000+1b
4000]
[ 7.964578] abrt-hook-ccpp[994]: segfault at 0 ip 00007fba2e09431b sp
00007ffefa92fb50 error 4 in libreport.so.0.0.1[7fba2e0860
00+25000]
[ 7.965398] Process 994(abrt-hook-ccpp) has RLIMIT_CORE set to 1
[ 7.965483] Aborting core
At mount time, The console displays these messages as well:
RPC: AUTH_GSS upcall timed out.
Please check user daemon is running.
gssd will start without issue if started manually on its own, but dies
if I subsequently try to manually mount any of the nfs mounts.
I've isolated the issue down to the latest update in the nfs-utils package:
Working: nfs-utils-1.3.0-0.21.el7.x86_64
Broken: nfs-utils-1.3.0-0.21.el7_2.x86_64
(note the '_v2' difference)
gssd/the mounting works without issue after backing down to the older
nfs-utils version.
It looks similar to the following bug report (fixed in 1.3.1), but I'm
not 100% convinced it's the same:
https://bugzilla.redhat.com/show_bug.cgi?id=1108615
If this is in fact the issue, I would be happy keeping the updated
nfs-utils version, but working around the issue via changing my
krb5.conf, but I'm not sure where the *preferred*_realm comes from - I
do have *default*_realm set in my krb5.conf.
If that's not it, I'd be happy to provide any additional information
that may assist in troubleshooting & welcome any suggestions, but I'd
greatly prefer to retain the OS-supplied nfs-utils.
-Mark
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
