SUMMARY: SMB mount timeout errors
YOUR DETAILED BUG DESCRIPTION: Samba shares SMB mounts timeout after a certain period of time. See next section for details.
HOW CAN THIS BUG BE REPRODUCED?
1. I have a Samba share defined in a SLES9 SP1 server. Here's the Samba
configuration file:
# Samba configuration file
[global]
socket options=TCP_NODELAY
os level=33
wins proxy=yes
config file=/etc/samba/smb.conf
encrypt passwords=yes
workgroup=linuxdev
security=SHARE
wins support=yes
share modes=yes
lm announce=yes
keepalive=30
[netbench]
guest ok = true
case sensitive=no
public=yes
mangle case=no
read only=no
writeable=yes
only guest = yes
printable = no
path=/home/smbshare
2. On a client running SLES9 SP1, I mount the "netbench" Samba share as an SMB share:
$ mount -t smbfs //<server-ip-addr>/netbench /mnt/iosmb -o guest
3. In the client, I start several instances of iozone (
www.iozone.org) against the SMB share, and run them continuosly for several hours:
$ iozone -RazMb /mnt/iosmb/file01.xls -f /mnt/iosmb/file01 -s 100m
$ iozone -RazMb /mnt/iosmb/file02.xls -f /mnt/iosmb/file02 -s 100m
$ iozone -RazMb /mnt/iosmb/file03.xls -f /mnt/iosmb/file03 -s 100m
...
4. After a while (hours, sometimes minutes), I get a LOT of error messages
and timeouts. Here's a sample of the errors messages in the client:
write: Input/output error
write: Input/output error
read: Input/output error
read: Input/output error
creat: Input/output error
creat: Input/output error
Here's what I get when I try to list my mounted filesystems in the client:
$ df -k
Filesystem 1K-blocks Used Available Use% Mounted on
LABEL=/ 35028620 8005124 27023496 23% /
tmpfs 128332 8 128324 1% /dev/shm
/dev/sda1 15521 7326 7394 50% /boot
df: `/mnt/iosmb': Input/output error
$
Here are the errors in /var/log/messages in the client:
Nov 16 14:00:11 linux kernel: smbfs: unsupported module, tainting kernel.
Nov 16 14:02:00 linux smbd[10296]: [2004/11/16 14:02:00, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 14:02:00 linux smbd[10296]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 14:02:02 linux smbd[10373]: [2004/11/16 14:02:02, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 14:02:02 linux smbd[10373]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 14:20:44 linux -- MARK --
Nov 16 14:38:10 linux smbd[22537]: [2004/11/16 14:38:10, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 14:38:10 linux smbd[22537]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 14:38:11 linux smbd[22537]: [2004/11/16 14:38:11, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 14:38:11 linux smbd[22537]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 14:38:11 linux smbd[22663]: [2004/11/16 14:38:11, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 14:38:11 linux smbd[22663]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 14:38:12 linux smbd[22663]: [2004/11/16 14:38:12, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 14:38:12 linux smbd[22663]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 14:44:07 linux kernel: smb_add_request: request [c2679180, mid=17804] timed out!
Nov 16 14:44:07 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 14:44:07 linux kernel: smb_add_request: request [c2679e80, mid=17805] timed out!
Nov 16 14:44:07 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 15:21:30 linux kernel: smb_add_request: request [c2679980, mid=36440] timed out!
Nov 16 15:21:30 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 15:21:30 linux kernel: smb_add_request: request [c2679480, mid=36441] timed out!
Nov 16 15:21:30 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 15:21:34 linux kernel: smb_file_write: //10.9.165.213-1400-iozonesmb12 validation failed, error=4294967291
Nov 16 15:55:06 linux smbd[1007]: [2004/11/16 15:55:06, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 15:55:06 linux smbd[1007]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 15:55:08 linux smbd[1024]: [2004/11/16 15:55:08, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 15:55:08 linux smbd[1024]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 15:55:08 linux smbd[1109]: [2004/11/16 15:55:08, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 15:55:08 linux smbd[1109]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 15:55:09 linux smbd[1109]: [2004/11/16 15:55:09, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 15:55:09 linux smbd[1109]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 15:57:28 linux kernel: smb_add_request: request [c2679c80, mid=59961] timed out!
Nov 16 15:57:28 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 15:57:28 linux kernel: smb_add_request: request [c2679980, mid=59962] timed out!
Nov 16 15:57:28 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 15:59:00 linux /USR/SBIN/CRON[25529]: (root) CMD ( rm - /var/spool/cron/lastrun/cron.hourly)
Nov 16 16:07:10 linux smbd[7536]: [2004/11/16 16:07:10, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 16:07:10 linux smbd[7536]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 16:07:11 linux smbd[7572]: [2004/11/16 16:07:11, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 16:07:11 linux smbd[7572]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 16:07:11 linux smbd[7724]: [2004/11/16 16:07:11, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 16:07:11 linux smbd[7724]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 16:07:12 linux smbd[7724]: [2004/11/16 16:07:12, 0] lib/util_sock.c:read_socket_data(367)
Nov 16 16:07:12 linux smbd[7724]: read_socket_data: recv failure for 4. Error = Connection reset by peer
Nov 16 16:08:11 linux smbd[15718]: [2004/11/16 16:08:11, 0] smbd/sesssetup.c:reply_sesssetup_and_X(656)
Nov 16 16:08:11 linux smbd[15718]: reply_sesssetup_and_X: Rejecting attempt at SPNEGO session setup when it was
not negoitiated.
Nov 16 16:37:09 linux kernel: smb_add_request: request [c2679d80, mid=10769] timed out!
Nov 16 16:37:09 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
Nov 16 16:37:09 linux kernel: smb_add_request: request [c2679080, mid=10770] timed out!
Nov 16 16:37:09 linux kernel: smb_writepage_sync: failed write, wsize=4096, write_ret=-5
...
5. In some cases, I can no longer access the SMB share from the client, and I have to reboot to fix the network hangup.
6. From Googling the error messages, it appears to be a Kernel 2.6 bug with SMB. Oddly enough, when I mount the Samba share as a CIFS mount the problem goes away. So it appears that a possible solution is to use CIFS instead of SMB, but I would still like to know why I get these network timeouts with SMB and my 2.6 Kernel (2.6.5-7.97) and whether there are any known bugs with SLES9 SP1 and SMB.
Has anyone ran into these problems??
