[bug] FC2/2.6.8.1 does not fallback to NFSv2

[bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

[ I tried to send this yesterday without actually subscribing, but it
has yet to make it past the list moderator.  I've given up and
subscribed to yet another mailing list to get this bug report out. ]

We have just brought up several Fedora Core 2 systems running the latest
2.6.8.1 kernel revision in our NIS/NFS+autofs environment.  The FC2/2.6
clients are refusing to mount one specific filesystem on our network.
This filesystem is exported by a Debian/woody system running a 2.4.18
kernel.  The kernel on this machine only supports NFSv2 exports.

When the FC2/2.6 client attempts to automount the filesystem on the 2.4
server we observe the following in the client's syslog:

  Aug 26 18:37:08 scylla automount[5056]: >> mount to NFS server 'reef'
  failed: server is down.
  Aug 26 18:37:08 scylla automount[5056]: mount(nfs): nfs: mount failure
  reef:/export/reef/httpd/ on /sb/www
  Aug 26 18:37:08 scylla automount[5056]: failed to mount /sb/www

And the following is observed in the server's syslog:

  Aug 26 18:37:12 reef kernel: svc: unknown version (3)

At this point the client gives up with the following error:

  [root@scylla root]# ls /sb/www
  ls: /sb/www: No such file or directory

None of our other Linux systems running 2.4 kernels have any trouble
mounting that filesystem.  They all automount it using NFSv2 as
expected.  All of our 2.4 systems also correctly mount using NFSv3 from
systems that support it.

Autofs is configured on the FC2 systems to permit NFSv2 mounts as shown
by the minproto option below:

  [root@scylla root]# mount | grep automount
  automount(pid1665) on /home type autofs (rw,fd=4,pgrp=1665,minproto=2,maxproto=4)
  automount(pid1695) on /sb type autofs (rw,fd=4,pgrp=1695,minproto=2,maxproto=4)
  automount(pid1739) on /scratch type autofs (rw,fd=4,pgrp=1739,minproto=2,maxproto=4)

The 2.6.8.1 kernel is configured with the following options:

  [root@scylla root]# grep -i nfs /usr/src/linux/.config
  CONFIG_NFS_FS=y
  CONFIG_NFS_V3=y
  # CONFIG_NFS_V4 is not set
  # CONFIG_NFS_DIRECTIO is not set
  CONFIG_NFSD=y
  CONFIG_NFSD_V3=y
  # CONFIG_NFSD_V4 is not set
  CONFIG_NFSD_TCP=y

I am able to manually mount this filesystem, but only if I explicitly
add the vers=2 option to the mount command:

  [root@scylla root]# mount -t nfs reef:/export/reef/httpd/ /mnt/temp
  mount to NFS server 'reef' failed: server is down.
  [root@scylla root]# mount -t nfs -o vers=2 reef:/export/reef/httpd/ /mnt/temp
  [root@scylla root]#

Given the fact that the mount fails manually until vers=2 is specified I
am inclined to believe this is a kernel NFS bug, not an autofs or mount
bug.

I would be glad to gather any additional information that might be
necessary to determine the cause of this behavior.  Can anyone reproduce
this or have advice on how I might fix it?

Thanks,

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  sunlight is the lowest frequency and therefore
                         is the biggest wave


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Trond Myklebust]

P=E5 fr , 27/08/2004 klokka 17:22, skreiv Brandon D. Valentine:

> Given the fact that the mount fails manually until vers=3D2 is specified =
I
> am inclined to believe this is a kernel NFS bug, not an autofs or mount
> bug.

Sorry. I conclude the exact opposite. The problem here is the "mount"
program autoprobe failing.

Cheers,
  Trond


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Trond Myklebust]

P=C3=A5 fr , 27/08/2004 klokka 17:36, skreiv Trond Myklebust:
> P=C3=A5 fr , 27/08/2004 klokka 17:22, skreiv Brandon D. Valentine:
>=20
> > Given the fact that the mount fails manually until vers=3D2 is specifie=
d I
> > am inclined to believe this is a kernel NFS bug, not an autofs or mount
> > bug.
>=20
> Sorry. I conclude the exact opposite. The problem here is the "mount"
> program autoprobe failing.

To be more specific, I believe the problem here is that you are using an
NFSv2-only kernel together with a version of mountd that is advertising
NFSv2 *and* NFSv3.

We used to have code in the kernel that retried using NFSv2 if the NFSv3
GETATTR call failed. I ripped that out of 2.6.x since people have had
ample time (I'd say 4-5 years is pretty ample) to correct their setups.
Add the '--no=E2=80=90nfs=E2=80=90version 3' flag to your rpc.mountd startu=
p script,
just as the mountd manpage clearly states you should do, and all will be
well.

Cheers,
  Trond


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

On Fri, Aug 27, 2004 at 05:51:30PM -0400, Trond Myklebust wrote:
> To be more specific, I believe the problem here is that you are using
> an NFSv2-only kernel together with a version of mountd that is
> advertising NFSv2 *and* NFSv3.

I would be very happy to find out that it was so simple, but it doesn't
appear to be.  The problem definitely appears to be on the client
system, not the server.

  [root@scylla root]# rpcinfo -u reef nfs
  program 100003 version 2 ready and waiting
  [root@scylla root]# rpcinfo -u reef nfs 3
  rpcinfo: RPC: Program/version mismatch; low version = 2, high version = 2
  program 100003 version 3 is not available

  reef:~# ps -Af | grep mountd
  root       395     1  0 Aug17 ?        00:00:00 /usr/sbin/rpc.mountd --no-nfs-version 3

> We used to have code in the kernel that retried using NFSv2 if the
> NFSv3 GETATTR call failed. I ripped that out of 2.6.x since people
> have had ample time (I'd say 4-5 years is pretty ample) to correct
> their setups.  Add the '--no???nfs???version 3' flag to your
> rpc.mountd startup script, just as the mountd manpage clearly states
> you should do, and all will be well.

I am willing to believe that this is not a kernel bug, but can you
suggest where else in the client userland I might look for the source of
this problem?  It seems that this must either be a kernel bug or a bug
in mount(8).

I am more than willing to gather any information you might need to
discover the source of the behavior I'm experiencing.  Attached is
strace(1) output for the mount(1) command.  Hopefully that'll make it
through mailman, but I'll leave you in the Cc in case it doesn't.  The
mount binary on this system is stripped, so I don't know how much more
useful debug information I can get out of it without compiling a debug
version of mount.  FWIW, the mount binary on this FC2 system comes from:

  [root@scylla root]# rpm -q --whatprovides /bin/mount
  util-linux-2.12-18

Let me know how I can help.

Thanks,

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  sunlight is the sun's weekly nonduplicating
                         shopper


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

[-- Attachment #1: Type: text/plain, Size: 275 bytes --]

And of course I didn't actually attach the strace output file so here it
comes.  Sorry for the extra message.

Thanks,

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  brandon is going for his counts

[-- Attachment #2: mount.strace --]
[-- Type: text/plain, Size: 12539 bytes --]

execve("/bin/mount", ["mount", "-t", "nfs", "reef:/export/reef/httpd/", "/mnt/temp/"], [/* 27 vars */]) = 0
uname({sysname="Linux", nodename="scylla.structbio.vanderbilt.edu", release="2.6.8.1", version="#5 SMP Wed Aug 25 16:09:17 CDT 2004", machine="i686"}) = 0
brk(0)                                  = 0x805c000
open("/etc/ld.so.preload", O_RDONLY)    = -1 ENOENT (No such file or directory)
open("/etc/ld.so.cache", O_RDONLY)      = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=4256539, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=144, st_size=71507, st_atime=2004/08/27-18:24:59, st_mtime=2004/08/26-10:03:26, st_ctime=2004/08/26-10:03:26}) = 0
old_mmap(NULL, 71507, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40000000
close(3)                                = 0
open("/lib/tls/libc.so.6", O_RDONLY)    = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\300\273"..., 512) = 512
fstat64(3, {st_dev=makedev(3, 2), st_ino=20972881, st_mode=S_IFREG|0755, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=2848, st_size=1455084, st_atime=2004/08/27-18:24:59, st_mtime=2004/05/11-08:20:12, st_ctime=2004/08/20-04:02:55}) = 0
old_mmap(0x747000, 1158124, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x747000
old_mmap(0x85c000, 16384, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x115000) = 0x85c000
old_mmap(0x860000, 7148, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x860000
close(3)                                = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40012000
mprotect(0x85c000, 8192, PROT_READ)     = 0
mprotect(0x743000, 4096, PROT_READ)     = 0
set_thread_area({entry_number:-1 -> 6, base_addr:0x40012460, limit:1048575, seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1, seg_not_present:0, useable:1}) = 0
munmap(0x40000000, 71507)               = 0
brk(0)                                  = 0x805c000
brk(0x807d000)                          = 0x807d000
brk(0)                                  = 0x807d000
open("/usr/lib/locale/locale-archive", O_RDONLY|O_LARGEFILE) = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=25166981, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=78568, st_size=40246896, st_atime=2004/08/27-18:22:19, st_mtime=2004/08/19-15:00:54, st_ctime=2004/08/19-15:00:54}) = 0
mmap2(NULL, 2097152, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40013000
close(3)                                = 0
umask(022)                              = 022
open("/dev/null", O_RDWR|O_LARGEFILE)   = 3
close(3)                                = 0
getuid32()                              = 0
geteuid32()                             = 0
lstat64("/etc/mtab", {st_dev=makedev(3, 2), st_ino=4256572, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=645, st_atime=2004/08/27-18:22:19, st_mtime=2004/08/27-18:09:34, st_ctime=2004/08/27-18:09:34}) = 0
stat64("/sbin/mount.nfs", 0xbffff730)   = -1 ENOENT (No such file or directory)
uname({sysname="Linux", nodename="scylla.structbio.vanderbilt.edu", release="2.6.8.1", version="#5 SMP Wed Aug 25 16:09:17 CDT 2004", machine="i686"}) = 0
gettimeofday({1093649099, 827523}, NULL) = 0
getpid()                                = 6904
open("/etc/resolv.conf", O_RDONLY)      = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=5146370, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=78, st_atime=2004/08/27-18:22:15, st_mtime=2004/08/19-15:17:43, st_ctime=2004/08/19-15:17:43}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40213000
read(3, "search structbio.vanderbilt.edu\n"..., 4096) = 78
read(3, "", 4096)                       = 0
close(3)                                = 0
munmap(0x40213000, 4096)                = 0
socket(PF_FILE, SOCK_STREAM, 0)         = 3
connect(3, {sa_family=AF_FILE, path="/var/run/nscd/socket"}, 110) = -1 ENOENT (No such file or directory)
close(3)                                = 0
open("/etc/nsswitch.conf", O_RDONLY)    = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=4195335, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=1718, st_atime=2004/08/27-18:22:19, st_mtime=2004/08/19-15:30:34, st_ctime=2004/08/19-15:30:34}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40213000
read(3, "#\n# /etc/nsswitch.conf\n#\n# An ex"..., 4096) = 1718
read(3, "", 4096)                       = 0
close(3)                                = 0
munmap(0x40213000, 4096)                = 0
open("/etc/ld.so.cache", O_RDONLY)      = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=4256539, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=144, st_size=71507, st_atime=2004/08/27-18:24:59, st_mtime=2004/08/26-10:03:26, st_ctime=2004/08/26-10:03:26}) = 0
old_mmap(NULL, 71507, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40213000
close(3)                                = 0
open("/lib/libnss_files.so.2", O_RDONLY) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0`\33\0\000"..., 512) = 512
fstat64(3, {st_dev=makedev(3, 2), st_ino=25167015, st_mode=S_IFREG|0755, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=104, st_size=50944, st_atime=2004/08/27-18:24:59, st_mtime=2004/05/11-08:20:08, st_ctime=2004/08/19-15:00:56}) = 0
old_mmap(NULL, 45724, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40225000
old_mmap(0x4022f000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x9000) = 0x4022f000
close(3)                                = 0
mprotect(0x4022f000, 4096, PROT_READ)   = 0
munmap(0x40213000, 71507)               = 0
open("/etc/host.conf", O_RDONLY)        = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=4194451, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=17, st_atime=2004/08/27-18:22:15, st_mtime=2000/07/23-10:06:54, st_ctime=2004/08/19-15:00:37}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40213000
read(3, "order hosts,bind\n", 4096)     = 17
read(3, "", 4096)                       = 0
close(3)                                = 0
munmap(0x40213000, 4096)                = 0
open("/etc/hosts", O_RDONLY)            = 3
fcntl64(3, F_GETFD)                     = 0
fcntl64(3, F_SETFD, FD_CLOEXEC)         = 0
fstat64(3, {st_dev=makedev(3, 2), st_ino=5146368, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=203, st_atime=2004/08/27-18:22:15, st_mtime=2004/08/19-15:17:43, st_ctime=2004/08/19-15:17:43}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40213000
read(3, "# Do not remove the following li"..., 4096) = 203
read(3, "", 4096)                       = 0
close(3)                                = 0
munmap(0x40213000, 4096)                = 0
open("/etc/ld.so.cache", O_RDONLY)      = 3
fstat64(3, {st_dev=makedev(3, 2), st_ino=4256539, st_mode=S_IFREG|0644, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=144, st_size=71507, st_atime=2004/08/27-18:24:59, st_mtime=2004/08/26-10:03:26, st_ctime=2004/08/26-10:03:26}) = 0
old_mmap(NULL, 71507, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40213000
close(3)                                = 0
open("/lib/libnss_nis.so.2", O_RDONLY)  = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\300\33"..., 512) = 512
fstat64(3, {st_dev=makedev(3, 2), st_ino=25167020, st_mode=S_IFREG|0755, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=88, st_size=43528, st_atime=2004/08/27-18:24:59, st_mtime=2004/05/11-08:20:08, st_ctime=2004/08/19-15:00:56}) = 0
old_mmap(NULL, 37440, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40231000
old_mmap(0x40239000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x7000) = 0x40239000
close(3)                                = 0
open("/lib/libnsl.so.1", O_RDONLY)      = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0pZq\000"..., 512) = 512
fstat64(3, {st_dev=makedev(3, 2), st_ino=25402408, st_mode=S_IFREG|0755, st_nlink=1, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=192, st_size=96248, st_atime=2004/08/27-18:24:59, st_mtime=2004/05/11-08:20:07, st_ctime=2004/08/20-04:03:03}) = 0
old_mmap(0x712000, 84192, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x712000
old_mmap(0x723000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x11000) = 0x723000
old_mmap(0x725000, 6368, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x725000
close(3)                                = 0
mprotect(0x723000, 4096, PROT_READ)     = 0
mprotect(0x40239000, 4096, PROT_READ)   = 0
munmap(0x40213000, 71507)               = 0
uname({sysname="Linux", nodename="scylla.structbio.vanderbilt.edu", release="2.6.8.1", version="#5 SMP Wed Aug 25 16:09:17 CDT 2004", machine="i686"}) = 0
open("/var/yp/binding/structbio.2", O_RDONLY) = 3
readv(3, [{"\377\377", 2}, {"\1\0\0\0\240\201\212\370\3U\0\0", 12}], 2) = 14
gettimeofday({1093649099, 830821}, NULL) = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_UDP) = 4
bind(4, {sa_family=AF_INET, sin_port=htons(720), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
ioctl(4, FIONBIO, [1])                  = 0
setsockopt(4, SOL_IP, IP_RECVERR, [1], 4) = 0
fcntl64(4, F_SETFD, FD_CLOEXEC)         = 0
close(3)                                = 0
sendto(4, "LH!\371\0\0\0\0\0\0\0\2\0\1\206\244\0\0\0\2\0\0\0\3\0\0"..., 80, 0, {sa_family=AF_INET, sin_port=htons(853), sin_addr=inet_addr("160.129.138.248")}, 16) = 80
poll([{fd=4, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
recvfrom(4, "LH!\371\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\1"..., 8800, 0, {sa_family=AF_INET, sin_port=htons(853), sin_addr=inet_addr("160.129.138.248")}, [16]) = 140
close(4)                                = 0
time(NULL)                              = 1093649099
socket(PF_INET, SOCK_STREAM, IPPROTO_TCP) = 3
bind(3, {sa_family=AF_INET, sin_port=htons(721), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
connect(3, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("160.129.138.217")}, 16) = 0
write(3, "\200\0\0008\10\217\226\357\0\0\0\0\0\0\0\2\0\1\206\240"..., 60) = 60
poll([{fd=3, events=POLLIN, revents=POLLIN}], 1, 60000) = 1
read(3, "\200\0\0\34\10\217\226\357\0\0\0\1\0\0\0\0\0\0\0\0\0\0"..., 400) = 32
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_UDP) = 3
bind(3, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
ioctl(3, FIONBIO, [1])                  = 0
setsockopt(3, SOL_IP, IP_RECVERR, [1], 4) = 0
sendto(3, "\16\203\177\r\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\2\0\0\0"..., 56, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("160.129.138.217")}, 16) = 56
poll([{fd=3, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
recvfrom(3, "\16\203\177\r\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 400, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("160.129.138.217")}, [16]) = 28
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_UDP) = 3
bind(3, {sa_family=AF_INET, sin_port=htons(723), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
ioctl(3, FIONBIO, [1])                  = 0
setsockopt(3, SOL_IP, IP_RECVERR, [1], 4) = 0
sendto(3, "7Ee\330\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\2\0\0\0\3\0\0"..., 56, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("160.129.138.217")}, 16) = 56
poll([{fd=3, events=POLLIN, revents=POLLIN}], 1, 5000) = 1
recvfrom(3, "7Ee\330\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\10"..., 400, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("160.129.138.217")}, [16]) = 28
close(3)                                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_UDP) = 3
bind(3, {sa_family=AF_INET, sin_port=htons(724), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
ioctl(3, FIONBIO, [1])                  = 0
setsockopt(3, SOL_IP, IP_RECVERR, [1], 4) = 0
sendto(3, "K\236#\v\0\0\0\0\0\0\0\2\0\1\206\243\0\0\0\3\0\0\0\0\0"..., 40, 0, {sa_family=AF_INET, sin_port=htons(2049), sin_addr=inet_addr("160.129.138.217")}, 16) = 40
poll([{fd=3, events=POLLIN, revents=POLLIN}], 1, 3000) = 1
recvfrom(3, "K\236#\v\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\2\0\0\0"..., 400, 0, {sa_family=AF_INET, sin_port=htons(2049), sin_addr=inet_addr("160.129.138.217")}, [16]) = 32
close(3)                                = 0
close(3)                                = -1 EBADF (Bad file descriptor)
write(2, "mount to NFS server \'reef\' faile"..., 33) = 33
write(2, ": server is down.\n", 18)     = 18
exit_group(32)                          = ?

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Trond Myklebust]

P=E5 fr , 27/08/2004 klokka 19:34, skreiv Brandon D. Valentine:
> And of course I didn't actually attach the strace output file so here it
> comes.  Sorry for the extra message.

It all looks OK, apart from the fact that the mount program gives up
after it receives the PROG_MISMATCH reply from the server.

Mind applying the following debugging line to the util-linux SRPM, then
using the resulting "mount" to test out what is going wrong?

Cheers,
  Trond


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

On Sat, Aug 28, 2004 at 01:35:05AM -0400, Trond Myklebust wrote:
> P? fr , 27/08/2004 klokka 19:34, skreiv Brandon D. Valentine:
> > And of course I didn't actually attach the strace output file so here it
> > comes.  Sorry for the extra message.
> 
> It all looks OK, apart from the fact that the mount program gives up
> after it receives the PROG_MISMATCH reply from the server.
> 
> Mind applying the following debugging line to the util-linux SRPM, then
> using the resulting "mount" to test out what is going wrong?

I would be happy to, but you forgot to include the debugging line.  =)

Thanks,

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  sunlight is a project of the california voter
                         foundation to advance electronic filing and
                         internet disclosure of campaign finance data


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Trond Myklebust]

[-- Attachment #1: Type: text/plain, Size: 224 bytes --]

På lau , 28/08/2004 klokka 02:29, skreiv Brandon D. Valentine:
> I would be happy to, but you forgot to include the debugging line.  =)

Urgh. Mental note to stop sending mail on Fridays at 1:30am...

Cheers,
  Trond

[-- Attachment #2: mount_debug.dif --]
[-- Type: text/plain, Size: 468 bytes --]

 nfsmount.c |    2 ++
 1 files changed, 2 insertions(+)

--- util-linux-2.12/mount/nfsmount.c.orig	2004-08-28 00:33:05.768456000 -0400
+++ util-linux-2.12/mount/nfsmount.c	2004-08-28 01:33:23.436105000 -0400
@@ -310,6 +310,8 @@
 			 TIMEOUT);
 	clnt_destroy(clnt);
 	close(sock);
+	printf("Probed program %lu, vers %lu, (%s) with result %d\n",
+			prog, vers, prot == IPPROTO_UDP ? "udp" : "tcp", stat)
 	if (stat != RPC_PROGVERSMISMATCH)
 		return 1;
 

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

On Sat, Aug 28, 2004 at 02:45:22PM -0400, Trond Myklebust wrote:
> P? lau , 28/08/2004 klokka 02:29, skreiv Brandon D. Valentine:
> > I would be happy to, but you forgot to include the debugging line.  =)
> 
> Urgh. Mental note to stop sending mail on Fridays at 1:30am...

I know how you feel.  FYI, for anyone else who wants to apply that
patch, it's missing a semicolon.  ;-)

I've applied this to the SRPM, built the mount binary, and got the
following:

  [root@scylla mount]# ./mount -t nfs reef:/export/reef/httpd/ /mnt/temp/
  Probed program 100003, vers 3, (udp) with result 9
  mount to NFS server 'reef' failed: server is down.

I applied the patch James Pearson also posted to this thread to the
SRPM, and it allowed me to mount the filesystem.

  [root@scylla mount]# ./mount -t nfs reef:/export/reef/httpd/ /mnt/temp/
  Probed program 100003, vers 3, (udp) with result 9
  Probed program 100003, vers 2, (udp) with result 0
  Probed program 100005, vers 1, (tcp) with result 0
  [root@scylla mount]# grep reef /proc/mounts
  reef:/export/reef/httpd/ /mnt/temp nfs rw,v2,rsize=8192,wsize=8192,hard,udp,lock,addr=reef 0 0

Out of curiosity I just grabbed the upstream util-linux-2.12b tarball
and built it and was also able to mount the filesystem no problem, so
this bug definitely seems to be a bug in the RedHat SRPM.  Their version
of nfsmount.c is vastly different from the upstream version.

  [root@scylla mount]# ./mount -t nfs reef:/export/reef/httpd/ /mnt/temp/
  [root@scylla mount]# grep reef /proc/mounts 
  reef:/export/reef/httpd/ /mnt/temp nfs rw,v2,rsize=8192,wsize=8192,hard,udp,lock,addr=reef 0 0

Thanks for all of your help.  I guess from here on out I've gotta take
it up with the RPM maintainer.

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  valentine is helping to unravel the mystery of
                         how sound is interpreted 


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[James Pearson]

Brandon D. Valentine wrote:
> On Sat, Aug 28, 2004 at 02:45:22PM -0400, Trond Myklebust wrote:
> 
> I applied the patch James Pearson also posted to this thread to the
> SRPM, and it allowed me to mount the filesystem.
> 
>   [root@scylla mount]# ./mount -t nfs reef:/export/reef/httpd/ /mnt/temp/
>   Probed program 100003, vers 3, (udp) with result 9
>   Probed program 100003, vers 2, (udp) with result 0
>   Probed program 100005, vers 1, (tcp) with result 0
>   [root@scylla mount]# grep reef /proc/mounts
>   reef:/export/reef/httpd/ /mnt/temp nfs rw,v2,rsize=8192,wsize=8192,hard,udp,lock,addr=reef 0 0
> 
> Out of curiosity I just grabbed the upstream util-linux-2.12b tarball
> and built it and was also able to mount the filesystem no problem, so
> this bug definitely seems to be a bug in the RedHat SRPM.  Their version
> of nfsmount.c is vastly different from the upstream version.

Yes, its definitely a bug in Redhat's version - a bugzilla comment says 
'Patch added to CVS for the next util-linux package' - as yet, this 
doesn't seem to have been released yet.

It appears that a lot of the changes in their nfsmount.c are based on a 
patch that has been on Trond's site for sometime - 
http://client.linux-nfs.org/util-linux-2.11n.dif.gz (although, not the 
bit that causes it to fail to fall back to NFSv2!)

> 
>   [root@scylla mount]# ./mount -t nfs reef:/export/reef/httpd/ /mnt/temp/
>   [root@scylla mount]# grep reef /proc/mounts 
>   reef:/export/reef/httpd/ /mnt/temp nfs rw,v2,rsize=8192,wsize=8192,hard,udp,lock,addr=reef 0 0
> 
> Thanks for all of your help.  I guess from here on out I've gotta take
> it up with the RPM maintainer.

May be someone on this list from RedHat can help ...

James Pearson



-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Steve Dickson]

Brandon D. Valentine wrote:

>On Fri, Aug 27, 2004 at 05:51:30PM -0400, Trond Myklebust wrote:
>  
>
>>To be more specific, I believe the problem here is that you are using
>>an NFSv2-only kernel together with a version of mountd that is
>>advertising NFSv2 *and* NFSv3.
>>    
>>
>
>I would be very happy to find out that it was so simple, but it doesn't
>appear to be.  The problem definitely appears to be on the client
>system, not the server.
>
>  [root@scylla root]# rpcinfo -u reef nfs
>  program 100003 version 2 ready and waiting
>  [root@scylla root]# rpcinfo -u reef nfs 3
>  rpcinfo: RPC: Program/version mismatch; low version = 2, high version = 2
>  program 100003 version 3 is not available
>
>  reef:~# ps -Af | grep mountd
>  root       395     1  0 Aug17 ?        00:00:00 /usr/sbin/rpc.mountd --no-nfs-version 3
>  
>
Upgrade your nfs-utils to the latest rawhide version (nfs-utils-1.0.6-31 
I believe)
to see if it take care of the problem. I think this has to do with the 
order of
how mountd and rpc.imapd are started. Basically mountd has to be running
before rpc.idmapd  for things to work...  So in  the case of the server 
(where
rpc.idmapd is already started), the nfs startup script send rpc.idmapd a 
signal
telling it to reopen its nfsd files (i.e. files under /proc/fs/nfsd )

SteveD.



-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

On Sat, Aug 28, 2004 at 01:10:14PM -0400, Steve Dickson wrote:
> Upgrade your nfs-utils to the latest rawhide version
> (nfs-utils-1.0.6-31 I believe) to see if it take care of the problem.
> I think this has to do with the order of how mountd and rpc.imapd are
> started. Basically mountd has to be running before rpc.idmapd  for
> things to work...  So in  the case of the server (where rpc.idmapd is
> already started), the nfs startup script send rpc.idmapd a signal
> telling it to reopen its nfsd files (i.e. files under /proc/fs/nfsd )

I do not see how rpc.idmapd has anything at all to do with NFSv2/3, it
is an NFSv4ism.  Also, rpc.mountd on the server is working fine, as
evidenced by the rpcinfo output I included in a previous message and the
fact that every other client on the network mounts filesystems from it
successfully.

In other words: what the heck are you talking about? =)

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  sunshine is a great game that anyone will enjoy


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Steve Dickson]

Brandon D. Valentine wrote

>In other words: what the heck are you talking about? =)
>  
>
 I thought  the problem was that rpc.mount was being
*unintentionally* started with the " -no-nfs-version  3"
flag which I was  happening in my testing of the nfs4 code
Something similar to:

root       395     1  0 Aug17 ?        00:00:00 /usr/sbin/rpc.mountd 
--no-nfs-version 3

But I guess I was mistaken......

SteveD.


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[Brandon D. Valentine]

On Sat, Aug 28, 2004 at 09:32:44PM -0400, Steve Dickson wrote:
> Brandon D. Valentine wrote
> >In other words: what the heck are you talking about? =)
> > 
> I thought  the problem was that rpc.mount was being
> *unintentionally* started with the " -no-nfs-version  3"
> flag which I was  happening in my testing of the nfs4 code
> Something similar to:
> 
> root       395     1  0 Aug17 ?        00:00:00 /usr/sbin/rpc.mountd 
> --no-nfs-version 3
> 
> But I guess I was mistaken......

No problem.  Thanks for taking the the time to respond.

BTW, I noticed you've worked on the util-linux package some recently.
Can you look into whether the patch James Pearson posted in this thread
will make it into Fedora Core 3 Test 2?  It would be a shame for Fedora
Core 3 to be released with a /bin/mount that doesn't work with NFSv2
servers, especially when the fix is so small.

Thanks,

Brandon D. Valentine
-- 
brandon@dvalentine.com                           http://www.geekpunk.net
Pseudo-Random Googlism:  summer is boom time for building


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[raven]

On Fri, 27 Aug 2004, Brandon D. Valentine wrote:

> Autofs is configured on the FC2 systems to permit NFSv2 mounts as shown
> by the minproto option below:
> 
>   [root@scylla root]# mount | grep automount
>   automount(pid1665) on /home type autofs (rw,fd=4,pgrp=1665,minproto=2,maxproto=4)
>   automount(pid1695) on /sb type autofs (rw,fd=4,pgrp=1695,minproto=2,maxproto=4)
>   automount(pid1739) on /scratch type autofs (rw,fd=4,pgrp=1739,minproto=2,maxproto=4)

The minproto and maxproto have nothing to to with NFS version. They are 
related to the autofs kernel module <-> automount daemon communication.

Ian



-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: [bug] FC2/2.6.8.1 does not fallback to NFSv2

[James Pearson]

It's a bug in Redhat's util-linux package - see:

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=123416

As yet, I haven't found a fixed version available from Redhat - my 
simple workround is as follows - I guess it's OK, as this check doesn't 
seem to exist in other versions of nfsmount.c

James Pearson

*** util-linux-2.12/mount/nfsmount.c.dist       2004-07-08 
15:08:38.000000000 +0
100
--- util-linux-2.12/mount/nfsmount.c    2004-07-08 15:15:28.000000000 +0100
***************
*** 423,430 ****
--- 423,435 ----
                                 return 1;
                         memcpy(mnt_pmap, &save_mnt, sizeof(*mnt_pmap));
                 }
+
+               /* the following check breaks the automatic fall back
+                  to NFSv2 - JCP 8/7/2004 */
+ #if 0
                 if (rpc_createerr.cf_stat != RPC_PROGNOTREGISTERED)
                         break;
+ #endif
                 memcpy(nfs_pmap, &save_nfs, sizeof(*nfs_pmap));
         }
    out_bad:

Brandon D. Valentine wrote:
> [ I tried to send this yesterday without actually subscribing, but it
> has yet to make it past the list moderator.  I've given up and
> subscribed to yet another mailing list to get this bug report out. ]
> 
> We have just brought up several Fedora Core 2 systems running the latest
> 2.6.8.1 kernel revision in our NIS/NFS+autofs environment.  The FC2/2.6
> clients are refusing to mount one specific filesystem on our network.
> This filesystem is exported by a Debian/woody system running a 2.4.18
> kernel.  The kernel on this machine only supports NFSv2 exports.
> 
> When the FC2/2.6 client attempts to automount the filesystem on the 2.4
> server we observe the following in the client's syslog:
> 
>   Aug 26 18:37:08 scylla automount[5056]: >> mount to NFS server 'reef'
>   failed: server is down.
>   Aug 26 18:37:08 scylla automount[5056]: mount(nfs): nfs: mount failure
>   reef:/export/reef/httpd/ on /sb/www
>   Aug 26 18:37:08 scylla automount[5056]: failed to mount /sb/www
> 
> And the following is observed in the server's syslog:
> 
>   Aug 26 18:37:12 reef kernel: svc: unknown version (3)
> 
> At this point the client gives up with the following error:
> 
>   [root@scylla root]# ls /sb/www
>   ls: /sb/www: No such file or directory
> 
> None of our other Linux systems running 2.4 kernels have any trouble
> mounting that filesystem.  They all automount it using NFSv2 as
> expected.  All of our 2.4 systems also correctly mount using NFSv3 from
> systems that support it.
> 
> Autofs is configured on the FC2 systems to permit NFSv2 mounts as shown
> by the minproto option below:
> 
>   [root@scylla root]# mount | grep automount
>   automount(pid1665) on /home type autofs (rw,fd=4,pgrp=1665,minproto=2,maxproto=4)
>   automount(pid1695) on /sb type autofs (rw,fd=4,pgrp=1695,minproto=2,maxproto=4)
>   automount(pid1739) on /scratch type autofs (rw,fd=4,pgrp=1739,minproto=2,maxproto=4)
> 
> The 2.6.8.1 kernel is configured with the following options:
> 
>   [root@scylla root]# grep -i nfs /usr/src/linux/.config
>   CONFIG_NFS_FS=y
>   CONFIG_NFS_V3=y
>   # CONFIG_NFS_V4 is not set
>   # CONFIG_NFS_DIRECTIO is not set
>   CONFIG_NFSD=y
>   CONFIG_NFSD_V3=y
>   # CONFIG_NFSD_V4 is not set
>   CONFIG_NFSD_TCP=y
> 
> I am able to manually mount this filesystem, but only if I explicitly
> add the vers=2 option to the mount command:
> 
>   [root@scylla root]# mount -t nfs reef:/export/reef/httpd/ /mnt/temp
>   mount to NFS server 'reef' failed: server is down.
>   [root@scylla root]# mount -t nfs -o vers=2 reef:/export/reef/httpd/ /mnt/temp
>   [root@scylla root]#
> 
> Given the fact that the mount fails manually until vers=2 is specified I
> am inclined to believe this is a kernel NFS bug, not an autofs or mount
> bug.
> 
> I would be glad to gather any additional information that might be
> necessary to determine the cause of this behavior.  Can anyone reproduce
> this or have advice on how I might fix it?
> 
> Thanks,
> 
> Brandon D. Valentine


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs