Bug 431722

Summary: can't mount nfs
Product: [openSUSE] openSUSE 11.1 Reporter: Felix Miata <mrmazda>
Component: NetworkAssignee: Neil Brown <nfbrown>
Status: RESOLVED FIXED QA Contact: E-mail List <qa-bugs>
Severity: Normal    
Priority: P5 - None CC: mmeeks
Version: Factory   
Target Milestone: ---   
Hardware: i686   
OS: Other   
Whiteboard:
Found By: --- Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Attachments: chkconfig --list output
lspci -v output
list of rpms containing the cause of this bug

Description Felix Miata 2008-10-02 19:08:50 UTC
Created attachment 243195 [details]
chkconfig --list output

fstab entry:
a-865:/pub	/nfs/a-865/pub	nfs	rw,noauto,nosuid,soft,rsize=8192,wsize=8192,nolocks 0 0

#rpm -qa | grep nfs
limal-nfs-server-1.5.0-1.12
limal-nfs-server-perl-1.5.0-1.12
nfs-client-1.1.3-5.2
nfs-kernel-server-1.1.3-5.2
nfsidmap-0.20-1.105
yast2-nfs-client-2.17.4-1.4
yast2-nfs-common-2.17.0-1.63
yast2-nfs-server-2.17.0-1.63

Result of attempting to mount:
mount.nfs: rpc.statd is not running but is required for remote locking.
mount.nfs: Either use '-o nolock' to keep locks local, or start statd.

Using kernel-pae-2.6.27-12.1.i586 on PIII 1.133GHz Dell GX150 (i815). Same result after executing rpc.statd at root prompt. Same result without nolocks in fstab entry. Still works fine booted to 11.0 on same machine.
Comment 1 Felix Miata 2008-10-02 19:09:24 UTC
Created attachment 243196 [details]
lspci -v output
Comment 2 Felix Miata 2008-10-02 20:26:04 UTC
A fresh update on a Sempron/NForce2 system duplicated this.
Comment 3 Felix Miata 2008-10-02 22:36:40 UTC
A fresh update on a P4/845G system replicated this.
Comment 4 Felix Miata 2008-10-05 23:46:51 UTC
Created attachment 243533 [details]
list of rpms containing the cause of this bug

Probably wasted my time, but on a 5th system not upgraded in several months, I did the following:
smart upgrade --yes rpm smart
smart upgrade --yes dbus\* gli\* hal kpartx mkinitrd \*openSUSE\* perl-Bootloader \*splash\* udev yast2-bootloader zypper
smart download kernel-pae kernel-pae-base
rpm -ivh kernel-pae kernel-pae-base
reboot
test nfs mounting (pass)
smart upgrade --yes z\* y\* x\* mozilla\*
smart upgrade --yes k\* f\* m\* s\* t\* u\* v\* w\* opera epiphany
smart upgrade --yes a\* b\* c\* d\* e\* g\*
test nfs mounting (pass)
smart upgrade --yes h\*
smart upgrade --yes i\*
smart upgrade --yes j\*
smart upgrade --yes o\*
smart upgrade --yes p\* q\*
smart upgrade --yes r\*
test nfs mounting (pass)
smart upgrade --yes l\*
test nfs mounting (fail)

I also did a Rawhide update on one of the same 5 Factory machines, and it now has this problem too.
Comment 5 Andreas Jaeger 2008-10-06 19:01:00 UTC
*** Bug 432612 has been marked as a duplicate of this bug. ***
Comment 6 Felix Miata 2008-10-07 04:57:25 UTC
After reading Bug 431542 I installed rpcbind. It didn't help, but I've been unable to figure out how to enable it. Both YaST2 (which lies, claiming success) and chkconfig fail to turn it on, evidenced by chkconfig --list.
Comment 7 Felix Miata 2008-10-07 12:27:06 UTC
I tried disabling ipv6 in YaST2 and rebooting, but it didn't help. I noticed too that on tty1 that nfsserver failed to start (idmapd mountd statdstartproc: exit status of parent of /usr/sbin/rpc.statd: 1), confirmed by showmount -e failure on another machine that should be able to access it.
Comment 8 Matthias Koenig 2008-10-07 12:39:14 UTC
Of course this does not work, IPv6 has to be disabled in nfs-utils at build time.
Comment 9 Matthias Koenig 2008-10-07 16:40:15 UTC

*** This bug has been marked as a duplicate of bug 431542 ***
Comment 10 Felix Miata 2008-11-14 02:34:03 UTC
I updated a PIII early & later a Sempron system from mirrors today using pae 2.6.27.5-2. The PIII still works, but this bug is back on the Sempron. I see during init "Starting Samba SMB daemon; Starting smartd; idmapd mountd statd nfsdCannot register service: RPC: Unable to receive: errno = Connection refused; Starting SSH daemon" and a long long long long long long pause before "Failed services in runlevel 3: nfsserver", with both .5-2 and .4-2 kernels.
Comment 11 Neil Brown 2008-11-14 03:03:49 UTC
This sounds like a new bug, so it is probably best to open a new
bugzilla and leave this one closed.  Things can get confusing
otherwise.

It sounds like neither portmap or rpcbind is running.  Please check
one is installed, report which, and see if it is running when the system
finally comes up.
Comment 12 Felix Miata 2008-11-15 02:08:43 UTC
Portmap not installed; rpcbind not enabled. Works with rpcbind enabled. No idea why it wasn't unless some update removed portmap after I had disabled all networking in runlevel 2, and, not knowing what rpcbind was at the time, I disabled it entirely. I updated two more systems today, and both had rpcbind enabled, as did the first system. So, no current bug -> worksforme.
Comment 13 Neil Brown 2008-11-15 07:13:33 UTC
OK, thanks.
Closing bug accordingly.
Comment 14 Felix Miata 2008-12-09 06:20:47 UTC
I just did an 11.0 to 11.1-RC1 upgrade. It replaced portmap with rpcbind, without turning rpcbind on. So, the problem still exists for portmap users as of RC1 unless the user knows to do insserv rpcbind or enable rpcbind in YaST*.
Comment 15 Matthias Koenig 2008-12-09 10:49:45 UTC
Yes, this is not fixed in RC1 since this is missing the following fixes for rpcbind:

-------------------------------------------------------------------
Mon Dec  1 12:29:50 CET 2008 - mkoenig@suse.de

- if upgrading from portmap:
  * we need to stop portmap and start rpcbind in %post since the
    restart on upgrade mechanism does not work here
  * dump portmap tables and set them again in rpcbind [bnc#446388]
- add pmap_set.c which is needed to set the old portmap tables
  in rpcbind

-------------------------------------------------------------------
Wed Nov 26 14:18:06 CET 2008 - mkoenig@suse.de

- specify runlevel in chkconfig check [bnc#444663] 


But it is fixed in the RC2/GM release.
Comment 16 Neil Brown 2008-12-10 23:13:09 UTC
marking as FIXED based on last comment.