Hi,
(I'm not subscribed to this list)
When I had a temporary need for more diskspace on a XEN VM running SLES10
(x86_64), I tried to mount an NFS filesystem to Linux, but I had problems:
1) When Solaris 9 (SPARC) was the server (server1), mount succeeded, but locking
failed:
Application (SuSE "build") said:
initializing rpm db...
warning: waiting for transaction lock on /var/lib/rpm/__db.000
error: can't create transaction lock on /var/lib/rpm/__db.000
/var/log/messages said:
Nov 16 10:27:31 nfs-client kernel: lockd: failed to monitor server1
Nov 16 10:27:31 nfs-client kernel: lockd/statd: failed to create
/var/lib/nfs/sm/server1: err=-2
2) When trying to use SLES9 (i386, Kernel 2.6.5-7.282-default) as NFS server
(server2), mount on the client hung hard (only kill -9 did work). The rpc.mountd
was waiting for data while the client hung:
strace:
open("/proc/fs/nfsd/filehandle", O_RDWR) = 10
fstat64(10, {st_mode=S_IFREG|0600, st_size=0, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
0x40019000
write(10, "*.dvm.our-domain,r"..., 85) = 85
read(10, "\\x0100000600fd000b020000006e2200"..., 4096) = 75
close(10) = 0
munmap(0x40019000, 4096) = 0
write(8, "\200\0\0Pb\376%\321\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 84) = 84
select(1024, [3 4 5 6 7 8], NULL, NULL, NULL) = 1 (in [8])
poll([{fd=8, events=POLLIN, revents=POLLIN}], 1, 35000) = 1
read(8, "", 4000) = 0
close(8) = 0
select(1024, [3 4 5 6 7], NULL, NULL, NULL <unfinished ...>
Tracing the network traffic, I found this final data exchange:
11:29:43.869678 IP (tos 0x0, ttl 64, id 50629, offset 0, flags [DF], proto: TCP
(6), length: 200) nfs-client.dvm.our-domain.2253585828 > server2.dvm.our-
domain.nfs: 148 fsinfo fh
Unknown/0100000600FD000B020000006E22000002000000000000000200000001000000
11:29:43.909121 IP (tos 0x0, ttl 64, id 29964, offset 0, flags [DF], proto: TCP
(6), length: 52) server2.dvm.our-domain.nfs > nfs-client.dvm.our-domain.1023: .,
cksum 0x655b (correct), ack 542460337 win 6432 <nop,nop,timestamp 470378359
19011966>
Any helpful instructions?
Note: There a some hints that the locking code in the NFS client may have a
problem, because I had a similar complaint for an NFS-mounted /home from HP-UX
11.11 (Shell hangs during login)
Regards,
Ulrich
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
On Thu, 2006-11-16 at 11:54 +0100, Ulrich Windl wrote:
> Hi,
>
> (I'm not subscribed to this list)
> When I had a temporary need for more diskspace on a XEN VM running SLES10
> (x86_64), I tried to mount an NFS filesystem to Linux, but I had problems:
>
> 1) When Solaris 9 (SPARC) was the server (server1), mount succeeded, but locking
> failed:
>
> Application (SuSE "build") said:
> initializing rpm db...
> warning: waiting for transaction lock on /var/lib/rpm/__db.000
> error: can't create transaction lock on /var/lib/rpm/__db.000
>
> /var/log/messages said:
> Nov 16 10:27:31 nfs-client kernel: lockd: failed to monitor server1
> Nov 16 10:27:31 nfs-client kernel: lockd/statd: failed to create
> /var/lib/nfs/sm/server1: err=-2
err == -2 is the same as ENOENT.
Does the directory /var/lib/nfs/sm exist and have the proper
permissions?
> 2) When trying to use SLES9 (i386, Kernel 2.6.5-7.282-default) as NFS server
> (server2), mount on the client hung hard (only kill -9 did work). The rpc.mountd
> was waiting for data while the client hung:
>
> strace:
> open("/proc/fs/nfsd/filehandle", O_RDWR) = 10
> fstat64(10, {st_mode=S_IFREG|0600, st_size=0, ...}) = 0
> mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
> 0x40019000
> write(10, "*.dvm.our-domain,r"..., 85) = 85
> read(10, "\\x0100000600fd000b020000006e2200"..., 4096) = 75
> close(10) = 0
> munmap(0x40019000, 4096) = 0
> write(8, "\200\0\0Pb\376%\321\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 84) = 84
> select(1024, [3 4 5 6 7 8], NULL, NULL, NULL) = 1 (in [8])
> poll([{fd=8, events=POLLIN, revents=POLLIN}], 1, 35000) = 1
> read(8, "", 4000) = 0
> close(8) = 0
> select(1024, [3 4 5 6 7], NULL, NULL, NULL <unfinished ...>
>
> Tracing the network traffic, I found this final data exchange:
>
> 11:29:43.869678 IP (tos 0x0, ttl 64, id 50629, offset 0, flags [DF], proto: TCP
> (6), length: 200) nfs-client.dvm.our-domain.2253585828 > server2.dvm.our-
> domain.nfs: 148 fsinfo fh
> Unknown/0100000600FD000B020000006E22000002000000000000000200000001000000
> 11:29:43.909121 IP (tos 0x0, ttl 64, id 29964, offset 0, flags [DF], proto: TCP
> (6), length: 52) server2.dvm.our-domain.nfs > nfs-client.dvm.our-domain.1023: .,
> cksum 0x655b (correct), ack 542460337 win 6432 <nop,nop,timestamp 470378359
> 19011966>
>
> Any helpful instructions?
No idea about this one. It looks as if your server is failing to reply
to the client's fsinfo inquiry. All it appears to have done is to ack
the TCP message.
Is mountd up and running?
Cheers
Trond
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
On 16 Nov 2006 at 12:10, Trond Myklebust wrote:
> On Thu, 2006-11-16 at 11:54 +0100, Ulrich Windl wrote:
> > Hi,
> >
> > (I'm not subscribed to this list)
> > When I had a temporary need for more diskspace on a XEN VM running SLES10
> > (x86_64), I tried to mount an NFS filesystem to Linux, but I had problems:
> >
> > 1) When Solaris 9 (SPARC) was the server (server1), mount succeeded, but locking
> > failed:
> >
> > Application (SuSE "build") said:
> > initializing rpm db...
> > warning: waiting for transaction lock on /var/lib/rpm/__db.000
> > error: can't create transaction lock on /var/lib/rpm/__db.000
> >
> > /var/log/messages said:
> > Nov 16 10:27:31 nfs-client kernel: lockd: failed to monitor server1
> > Nov 16 10:27:31 nfs-client kernel: lockd/statd: failed to create
> > /var/lib/nfs/sm/server1: err=-2
>
> err == -2 is the same as ENOENT.
>
> Does the directory /var/lib/nfs/sm exist and have the proper
> permissions?
Yes (I think):
Client: drwx------ 2 root root 48 Jul 3 21:11 /var/lib/nfs/sm
Server: drwx------ 2 root root 48 Nov 17 2005 /var/lib/nfs/sm
The odd thing is that when trying to _create_ a file in an existing directory,
ENOENT is probably an unexpected error message. Isn't it?
>
> > 2) When trying to use SLES9 (i386, Kernel 2.6.5-7.282-default) as NFS server
> > (server2), mount on the client hung hard (only kill -9 did work). The rpc.mountd
> > was waiting for data while the client hung:
> >
> > strace:
> > open("/proc/fs/nfsd/filehandle", O_RDWR) = 10
> > fstat64(10, {st_mode=S_IFREG|0600, st_size=0, ...}) = 0
> > mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
> > 0x40019000
> > write(10, "*.dvm.our-domain,r"..., 85) = 85
> > read(10, "\\x0100000600fd000b020000006e2200"..., 4096) = 75
> > close(10) = 0
> > munmap(0x40019000, 4096) = 0
> > write(8, "\200\0\0Pb\376%\321\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 84) = 84
> > select(1024, [3 4 5 6 7 8], NULL, NULL, NULL) = 1 (in [8])
> > poll([{fd=8, events=POLLIN, revents=POLLIN}], 1, 35000) = 1
> > read(8, "", 4000) = 0
> > close(8) = 0
> > select(1024, [3 4 5 6 7], NULL, NULL, NULL <unfinished ...>
> >
> > Tracing the network traffic, I found this final data exchange:
> >
> > 11:29:43.869678 IP (tos 0x0, ttl 64, id 50629, offset 0, flags [DF], proto: TCP
> > (6), length: 200) nfs-client.dvm.our-domain.2253585828 > server2.dvm.our-
> > domain.nfs: 148 fsinfo fh
> > Unknown/0100000600FD000B020000006E22000002000000000000000200000001000000
> > 11:29:43.909121 IP (tos 0x0, ttl 64, id 29964, offset 0, flags [DF], proto: TCP
> > (6), length: 52) server2.dvm.our-domain.nfs > nfs-client.dvm.our-domain.1023: .,
> > cksum 0x655b (correct), ack 542460337 win 6432 <nop,nop,timestamp 470378359
> > 19011966>
> >
> > Any helpful instructions?
>
> No idea about this one. It looks as if your server is failing to reply
> to the client's fsinfo inquiry. All it appears to have done is to ack
> the TCP message.
>
> Is mountd up and running?
The mountd was the process being strace'd, so I guess, yes, it is running.
Regards,
Ulrich
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs