openSUSE 11.3 NFS problem (hang) during boot process

English Install/Boot/Login
1

I’ve upgraded my openSUSE 11.2 to openSUSE 11.3.
Everything went great and i didn’t had any (dependency) problems during upgrade.

Although, there is one thing that got my attention. If NFS is enabled during boot, the system hangs for several minutes. Doesn’t seem to do anything since there’s no harddisk activity. Suddenly it continues with starting haldaemon (yes, i use KDE).

I’ve read the whole most annoying bug-list. And it appears there has been a problem with NFS during development cycle. Yet, it is marked as resolved.

Anyone an idea why I might have this problem still? I’ve tried recreating my nfs config since it seems to be changed a little in comparison to 11.2 but still it makes boot process hang if enabled.

Maybe this bug shouldn’t be marked as closed after all? I’m just very curious if anyone else is having this issue, because the release is very close from now.

2

I’ve upgraded from 11.2 to 11.3 today and have exactly the same problem. Seems like the bug is still alive!

3

Same problem for me
I just stopped nfs on system start to avoid the problem

4

Thanks for sharing.

Wondering if it’s the upgrade to blame?
Someone having this issue from fresh install?

5

Could it be the stale NFS locks problem? NFS sharing prior to reboot (my case crash) and reboot can’t find those locks on reboot. Stopping NFS fixes the problem. I don’t remember deleting any NFS locks. Also, are NFS versions for 11.2 and 11.3 the same?

6

seen this one ?

4.10. NFS Client and Server

…
From openSUSE 11.3 (SLE12 respectively) on, the structure of NFS client configuration has changed.
…
Attention: the old structure is not compatible with the new one…
…

7

Yeah, i’ve read that. Therefor, i have uninstalled nfs packages, get rid of all folders that contains NFS data/config, reïnstalled NFS and reconfigured my config.

Yet, when i enable NFS during boot, i still get the same error. Hang during boot process for several minutes.
Something tells me it’s more a idmapd problem than pure NFS since the boot message sequence stops there.
Unfortunately… i have no knowledge of idmapd so i can’t try to fix that very soon. (I’m reading docs now about that)

8

Upgrading to nfs-client 1.2.1-24.5 (Factory) doesn’t helps…

9

On 07/18/2010 11:26 AM, openLenny wrote:
>
> Upgrading to nfs-client 1.2.1-24.5 (Factory) doesn’t helps…

NFS is started by /etc/init.d/rc5.d/S05nfs. As the links in this directory are
executed alphabetically, this is after S02network, but well before S10xdm, which
will start the graphics manager. As wireless is not started until after the
desktop is up, you should be able to minimize the delay by


sudo mv /etc/init.d/rc5.d/S05nfs /etc/init.d/rc5.d/S99nfs

That way NFS will not be started until the last possible moment.

10

I do! I’m reaching my NAS via NFS and did so happily until I recently upgraded from 11.2 to 11.3.
Not only a fresh new installation on my netbook had this issue for I had the same experience with my desktop, which I upgraded. Now there’s a # in my fstabs! Hope this get’s fixed very soon.

11

Elaboration:

Even more strange is the fact I cannot reach my NAS via a nfs-mount command in the console!

mount.nfs //192.168.1.11:/mnt/IDE1/public /media/nas/
mount.nfs: DNS resolution failed for //192.168.1.11: Name or service not known

Via samba and a cifs-mount there’s no problems.

So the problem is not that I cannot access my data on my NAS but I find it very strange that I cannot reach it via NFS. No settings on my NAS have changed.
The NAS is an ICY BOX NAS 2000 with Firmware Version 2.3.2-mu-02.2, the NFS-server settings are as follows on this device:

root@nasbox:/etc # cat /etc/exports 
#
# the next lines will export your entire filesystem.
#
# DONT DO THIS UNTIL YOU REALLY KNOW WHAT YOU ARE DOING
#
#/ *(rw,no_root_squash,no_all_squash,async)
#/mnt/IDE1 *(rw,no_root_squash,no_all_squash,async)
#/system *(rw,no_root_squash,no_all_squash,async)
#
# entries added by sausalitos
/mnt/IDE1/public *(rw,root_squash)
12

That syntax doesn’t look right. It should be:

mount.nfs 192.168.1.11:/mnt/IDE1/public /media/nas/

i.e., without the double slash before the server name.

13

Thanks, you’re right! This, however gives:


mount.nfs 192.168.1.11:/mnt/IDE1/public /media/nas/
mount.nfs: Connection timed out
14

I tried your tip but I don’t have any network until nfs has started.
The only solution I found is to switch off nfs at start (ie : “chkconfig nfs off” as root) and start it manually after boot

15

I had the same symptom with mount.nfs timing out. Try using -v after mount.nfs and I think you will see it defaults to nfs4 and if your server does not support this it times out after several attempts. Edit /etc/nfsmount.conf as root and find the line with #Defaultvers=4, remove the # and change 4 to 3. That worked for me at least.

16

Thanks, dochm! That worked!

**mount.nfs  192.168.1.11:/mnt/IDE1/public /media/nas/ -v**
mount.nfs: timeout set for Wed Jul 21 21:05:55 2010
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=4,clientaddr=192.168.1.55'
mount.nfs: mount(2): Connection refused
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=4,clientaddr=192.168.1.55'
mount.nfs: mount(2): Connection refused
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=4,clientaddr=192.168.1.55'
mount.nfs: mount(2): Connection refused                                                                                                                         
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=4,clientaddr=192.168.1.55'                                                                         
mount.nfs: mount(2): Connection refused                                                                                                                         
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=4,clientaddr=192.168.1.55'                                                                         
mount.nfs: mount(2): Connection refused                                                                                                                         
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=4,clientaddr=192.168.1.55'                                                                         
^C

After changing the Defaultversion to 3 in/etc/nfsmount.conf:

**mount.nfs  192.168.1.11:/mnt/IDE1/public /media/nas/ -v  **                                                                                       
mount.nfs: timeout set for Wed Jul 21 21:07:41 2010                                                                                                               
mount.nfs: prog 100003, trying vers=3, prot=6                                                                                                                     
mount.nfs: portmap query retrying: RPC: Program not registered                                                                                                    
mount.nfs: prog 100003, trying vers=3, prot=17                                                                                                                    
mount.nfs: trying 192.168.1.11 prog 100003 vers 3 prot UDP port 2049                                                                                              
mount.nfs: prog 100005, trying vers=3, prot=17                                                                                                                    
mount.nfs: trying 192.168.1.11 prog 100005 vers 3 prot UDP port 725                                                                                               
mount.nfs: trying text-based options 'addr=192.168.1.11,vers=3,proto=udp,mountvers=3,mountproto=udp,mountport=725'                                                
192.168.1.11:/mnt/IDE1/public on /media/nas type nfs (nfsvers=3)

THANKS!!!

17

“Edit /etc/nfsmount.conf as root and find the line with #Defaultvers=4, remove the # and change 4 to 3.”

Thanks a ton man… that did the trick for me also! No more stalling in boot process with NFS enabled after making this change.
Weird actually, since op openSUSE 11.2 i always worked with NFSv4 enabled and never had any problems.

This should get a fix via online update soon! Maybe something that detects wich version the server supports and sets the config according to it (3 or 4) ? I’m not a programmer so i don’t know if that would be possible.

18

thanks, that did the trick for me, too…

but: why does this only occur, if networkmanager is enabled (and, thus, at nfs startup there is not yet an active network connection, because networkmanager is started later)?

if i disable networkmanager, everything works fine (eth0 is up and running).

if i enable networkmanager, idmapd hangs (until i change Defaultvers to 3).

can anyone enlighten me?

19

On 2010-07-18 19:26, Larry Finger wrote:

>


> sudo mv /etc/init.d/rc5.d/S05nfs /etc/init.d/rc5.d/S99nfs
>

NEVER, EVER do that on a SuSE/openSUSE/SLED/SLES system.

Read man init.d and do it the suse way, or you will break things.

–
Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” GM (Elessar))

20

Oh boys and girls, am I glad I know these forums. Had this trouble on a laptop, knew there was something about this in the forums, took the laptop home, changed to Defaultvers=3, problem solved. I love it.

Thanks to all !!