| View previous topic :: View next topic |
| Author |
Message |
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
 Posted: Wed Sep 15, 2004 12:39 pm Post subject: NFS export problem [RESOLVED] |
 |
|
I recently emerged the new version of dhcpcd which went into x86 stable. However, when I restarted all the network services I got an error when exporting NFS directories: | Code: | * Stopping NFS mountd... [ ok ]
* Stopping NFS daemon... [ ok ]
* Stopping NFS statd... [ ok ]
* Starting NFS statd... [ ok ]
* Exporting NFS directories...
/sbin/runscript.sh: line 534: 9259 Killed $exportfs -r 1>&2
* Error exporting NFS directories [ !! ]
* Starting NFS daemon... [ ok ]
* Starting NFS mountd... [ ok ] |
Needless to say, none of the directories are actually exported and no one can mount them. Line 534 of /sbin/runscript.sh is the very last line of the script, so that didn't help much.
I looked in /var/log/messages, and this error was reported: | Code: | Sep 15 13:26:36 Commodore64 rpc.mountd: Caught signal 15, un-registering and exiting.
Sep 15 13:26:36 Commodore64 nfsd: last server has exited
Sep 15 13:26:36 Commodore64 nfsd: unexporting all filesystems
Sep 15 13:26:36 Commodore64 rpc.statd[8850]: Caught signal 15, un-registering and exiting.
Sep 15 13:26:36 Commodore64 rpc.statd[9256]: Version 1.0.6 Starting
Sep 15 13:26:36 Commodore64 rpc.statd[9256]: gethostbyname error for Commodore64
Sep 15 13:27:06 Commodore64 rc-scripts: Error exporting NFS directories
Sep 15 13:27:06 Commodore64 svc: bad direction 65536, dropping request
Sep 15 13:27:06 Commodore64 svc: bad direction 65536, dropping request
|
I did a google search on "gethostbyname error" but nothing interesting turned up. Before I try downgrading dhcpcd, does anyone have any ideas on how I get this up and running again?
Last edited by chunderbunny on Thu Sep 23, 2004 12:17 pm; edited 2 times in total |
|
| Back to top |
|
 |
Nossie
Apprentice
Joined: 19 Apr 2002
Posts: 181
|
| Posted: Wed Sep 15, 2004 12:59 pm Post subject: |
|
|
it seems your system map it's own name to a ipnumber with the funcion 'gethostbyname'
Do you have an entry in /etc/hosts ?
[edit]
ehh, i meant /etc/hosts instead of /etc/host 
[/edit]
Last edited by Nossie on Wed Sep 15, 2004 1:57 pm; edited 1 time in total |
|
| Back to top |
|
|
hecatomb
Guru
Joined: 02 Sep 2004
Posts: 525
|
| Posted: Wed Sep 15, 2004 1:08 pm Post subject: |
|
|
Maybe you have not recieved a new ip-adress after the dhcpd update.
| Code: | | /etc/init.d/net.eth0 restart |
|
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Wed Sep 15, 2004 1:12 pm Post subject: |
|
|
Nope, /etc/host is empty. What is supposed to be in there?
And yes, I am getting an IP address. All other network services are up and running (ntp, ssh, etc.)
Edit: BTW, in case you meant /etc/hosts, that looksl ike this
| Code: | 127.0.0.1 localhost
# IPV6 versions of localhost and co
::1 ip6-localhost ip6-loopback
fe00::0 ip6-localnet
ff00::0 ip6-mcastprefix
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
ff02::3 ip6-allhosts |
And /etc/hostname is |
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Wed Sep 15, 2004 1:23 pm Post subject: |
|
|
Ok, I changed my /etc/hosts to this: | Code: | 127.0.0.1 Commodore64.retro Commodore64 localhost
# IPV6 versions of localhost and co
::1 ip6-localhost ip6-loopback
fe00::0 ip6-localnet
ff00::0 ip6-mcastprefix
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
ff02::3 ip6-allhosts |
And now the "gethostsbyname error" message doesn't appear when i try and restart NFS, but... it still doesn't export any directories. I'm still getting the error | Code: | /sbin/runscript.sh: line 534: 10059 Killed $exportfs -r 1>&2
* Error exporting NFS directories [ !! ] |
And the message | Code: | Sep 15 14:18:31 Commodore64 rpc.mountd: Caught signal 15, un-registering and exiting.
Sep 15 14:18:31 Commodore64 nfsd: last server has exited
Sep 15 14:18:31 Commodore64 nfsd: unexporting all filesystems
Sep 15 14:18:31 Commodore64 rpc.statd[9812]: Caught signal 15, un-registering and exiting.
Sep 15 14:18:31 Commodore64 rpc.statd[10056]: Version 1.0.6 Starting
Sep 15 14:19:01 Commodore64 rc-scripts: Error exporting NFS directories
Sep 15 14:19:02 Commodore64 svc: bad direction 65536, dropping request
|
Anything else? |
|
| Back to top |
|
|
hecatomb
Guru
Joined: 02 Sep 2004
Posts: 525
|
| Posted: Wed Sep 15, 2004 1:35 pm Post subject: |
|
|
I found this in the startscript:
| Code: | ...
# Exportfs likes to hang if networking isn't working.
# If that's the case, then try to kill it so the
# bootup process can continue.
... |
Are you sure your network connection is ok? ping some other hosts to test this.
Maybe the NFS servers hostname is also missing in your /etc/hosts? |
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Wed Sep 15, 2004 1:45 pm Post subject: |
|
|
Networking is definitely working, I'm administering this machine through an SSH connection. 
The machine with the trouble IS the NFS server, it's not exporting directories. I don't need a special entry in /etc/hosts for this.
Edit: Ok, this is interesting. If I change my /etc/exports so that I'm only exporting one directory, then it all works fine. With two entries in /etc/exports then it fails. It doesn't matter which entry I use either... |
|
| Back to top |
|
|
hecatomb
Guru
Joined: 02 Sep 2004
Posts: 525
|
| Posted: Wed Sep 15, 2004 1:59 pm Post subject: |
|
|
Sorry, didn't get the point (Commodore64 IS the NFS Server)!
This is strange, exporting one dir is ok but two fail. No idea! |
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Wed Sep 15, 2004 2:50 pm Post subject: |
|
|
| I upgraded nfs-utils to ~x86, but that hssn't changed anything. I can still only export one NFS directory at a time. |
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Thu Sep 16, 2004 11:01 pm Post subject: |
|
|
| Bump, I'm still not getting anywhere. Only 1 directory can be exported at a time without NFS throwing up an error. |
|
| Back to top |
|
|
hecatomb
Guru
Joined: 02 Sep 2004
Posts: 525
|
| Posted: Fri Sep 17, 2004 7:06 am Post subject: |
|
|
Maybe the output of exportfs will help.
I have no NFS so this is just a guess. The daemon will be started with a script like every other too. Find the script (e.g.)
| Code: | | find /etc -exec grep "Error exporting NFS directories" {} /dev/null \; |
Then make a backup and edit this file:
| Code: | ...
ebegin "Exporting NFS directories"
$exportfs -r 1>&2 &
pid=$!
( sleep 30; kill -9 $pid &>/dev/null ) &
wait $pid
eend $? "Error exporting NFS directories"
... |
edit these lines so that the command exportfs will NOT be killed after 30 seconds:
| Code: | ...
ebegin "Exporting NFS directories"
$exportfs -r
# pid=$!
# ( sleep 30; kill -9 $pid &>/dev/null ) &
# wait $pid
eend $? "Error exporting NFS directories"
... |
Or maybe you can execute the command by yourself.
| Code: | | (/usr/sbin/)exportfs -r |
Hope you will get some interesting messages! |
|
| Back to top |
|
|
barnie
n00b
Joined: 04 Jul 2003
Posts: 21
|
| Posted: Wed Sep 22, 2004 11:45 pm Post subject: Solved |
|
|
| I had the same problem. Running exportfs -r sepearte solved it. So I increased the time in the script. Now everything is working. |
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Wed Sep 22, 2004 11:50 pm Post subject: |
|
|
| Which script did you edit to increase the time? |
|
| Back to top |
|
|
sf_alpha
Tux's lil' helper
Joined: 19 Sep 2002
Posts: 136
Location: Bangkok, TH
|
| Posted: Thu Sep 23, 2004 2:52 am Post subject: |
|
|
May be portmap problems ... Checkit ?
_________________
Gentoo Mirrors in Thailand (and AP)
http://gentoo.in.th |
|
| Back to top |
|
|
hecatomb
Guru
Joined: 02 Sep 2004
Posts: 525
|
| Posted: Thu Sep 23, 2004 5:45 am Post subject: |
|
|
| Quote: | | Which script did you edit to increase the time? |
I think he means the startscript /etc/init.d/nfs
| Code: | ...
ebegin "Exporting NFS directories"
$exportfs -r 1>&2 &
pid=$!
( sleep 30; kill -9 $pid &>/dev/null ) &
wait $pid
eend $? "Error exporting NFS directories"
... |
Increase the 30 sec. the script is waiting for the end of exportfs. Maybe this will help! |
|
| Back to top |
|
|
chunderbunny
Veteran
Joined: 31 May 2004
Posts: 1281
Location: 51°24'27" N, 0°57'15" W
|
| Posted: Thu Sep 23, 2004 12:17 pm Post subject: |
|
|
Cool, thanks that worked! 
Well, not exactly, the script is a little different in my version (it has a seperate waitfor_exportfs function) but the general principle is the same. |
|
| Back to top |
|
|
richardash1981
Tux's lil' helper
Joined: 08 Apr 2005
Posts: 94
Location: England
|
| Posted: Fri Nov 11, 2005 10:18 pm Post subject: |
|
|
Your problem is waiting for network name resolution for the _clients_ to time out. You must be able to reverse DNS each client from the server (try running dig there).
The nfs server needs to be able to do a reverse DNS lookup on each client IP that tries to connect to it,(even with IP addresses or masks in /etc/exports). If it can't, mount will often time out on the clients, and exportfs will take much longer.
Either add the nfs clients to the DNS or the server's hosts file. |
|
| Back to top |
|
|
|
Networking & Security
