Samba: smb_add_request: request timed out!

I have a very strange problem with Samba (that might be a bug) but I'm looking for any work arounds that someone might know. Set-up, FC 2 updated last week and Windows 2k server.

# rpm -q samba
samba-3.0.10-1.fc2
# rpm -q kernel
kernel-2.6.5-1.358
kernel-2.6.10-1.8_FC2
# uname -a
Linux tavarua 2.6.10-1.8_FC2 #1 Mon Jan 10 14:46:18 EST 2005 i686 i686 i386 GNU/Linux

Here is what I am trying to do: Share a Windows 2k folder via Samba and then create a symbolic link to serve out from Apache.

# pwd;ls -al
/var/www/html/tavarua
lrwxrwxrwx  1 root root   20 Dec 31 00:30 cloudbreak -> /mnt/cloudbreak/Web/

I believe that Samba is configured correctly (I can mount my home directory from windows and my shares are mounted correctly on FC2), see my mount output:

//cloudbreak/My_Documents on /mnt/cloudbreak type smbfs (0)

Apache seems to be working fine for other direcotries.

My services smb, httpd, are running. I've restarted them and they appear to not resolve the issue. I keep getting error messages in my logs like the following:

Jan 22 15:27:51 tavarua nmbd[16475]: [2005/01/22 15:27:51, 0] nmbd/nmbd_become_lmb.c:become_local_master_stage2(396)
Jan 22 15:27:51 tavarua nmbd[16475]:   *****
Jan 22 15:27:51 tavarua nmbd[16475]:
Jan 22 15:27:51 tavarua nmbd[16475]:   Samba name server TAVARUA is now a local master browser for workgroup HAWAII on subnet 10.0.0.15
Jan 22 15:27:51 tavarua nmbd[16475]:
Jan 22 15:27:51 tavarua nmbd[16475]:   *****
Jan 22 15:28:05 tavarua mount.smbfs[2800]: [2005/01/22 15:28:05, 0] tdb/tdbutil.c:tdb_log(725)
Jan 22 15:28:05 tavarua mount.smbfs[2800]:   tdb(/var/cache/samba/gencache.tdb): tdb_lock failed on list 39 ltype=0 (Bad file descriptor)
Jan 22 15:28:05 tavarua mount.smbfs[2800]: [2005/01/22 15:28:05, 0] tdb/tdbutil.c:tdb_log(725)
Jan 22 15:28:05 tavarua mount.smbfs[2800]:   tdb(/var/cache/samba/gencache.tdb): tdb_lock failed on list 39 ltype=1 (Bad file descriptor)
Jan 22 15:28:05 tavarua mount.smbfs[2800]: [2005/01/22 15:28:05, 0] tdb/tdbutil.c:tdb_log(725)
Jan 22 15:28:05 tavarua mount.smbfs[2800]:   tdb(/var/cache/samba/gencache.tdb): tdb_lock failed on list 39 ltype=1 (Bad file descriptor)
Jan 22 15:28:05 tavarua kernel: SMB connection re-established (-5)
Jan 22 15:28:35 tavarua kernel: smb_add_request: request [c5d73ee0, mid=2400] timed out!
Jan 22 15:28:40 tavarua kernel: smb_add_request: request [c5d73dc0, mid=2401] timed out!
Jan 22 15:29:10 tavarua kernel: smb_add_request: request [c5d73dc0, mid=2402] timed out!
Jan 22 15:29:40 tavarua kernel: smb_add_request: request [c5d73dc0, mid=2403] timed out!
Jan 22 15:30:10 tavarua kernel: smb_add_request: request [c5d73dc0, mid=2404] timed out!

Does anyone know what these error messages are and possibly any workarounds. I think I have all the information needed to troubleshoot this but any help is appreciated.

A reboot of the FC 2 machine seems to clear this issue which leads me to believe that this is problem du eto cache or some sort of system build-up. Next time this occurs, I will see if rebooting the windows system has a similar effect.

Still, any knowledge on this problem is helpful to me. Thanks!

bttt

ive been having the same errors come up recently. have you found any info about it? ive been sifting through google but i think im looking in the wrong places.

I have found no solution. I think that this is an NFS issue thought, not a SMB issue. I've had success running:

service nfs restart

sometimes. (I think that's it) The errors, looking more closely, are coming from the kernel.

This is just annyoing.

I am having this same problem here @ work using unattended.sourceforge.net

One system here just refused to run the install, constantly getting this error even after several samba restarts on the server side.

I hope someone has some lighto n this topic.

I have looked around, but have found no answer to this original post. Anybody have any luck correcting this error?

It does seem to affect NFS as well as SMB shares. I am having a major problem on a new FC3 install w/3.0.10 samba.

I am seeing a ton of tdb_lock failed messages as well as getpeername failed. When it happens I see a long list of SMB connection re-established lines in dmesg.

I can't figure out where this is coming from and is driving me crazy as I have to reboot the box in order to release dead connection sockets. It doesn't help to kill the connections on the Windows boxes or on the other Linux boxes with the nfs shares.

I am running this on a Dell 2650. I saw a few posts with references to these boxes causing problems with the 2.6 kernel, but I don't want to believe that is it because I have 2 of these with the 2.6 kernel and the other one is flawless but not running samba or nfs.

kernel is Linux ftp1 2.6.10-1.766_FC3smp #1 SMP Wed Feb 9 23:21:37 EST 2005 i686 i686 i386 GNU/Linux


Any ideas yet??? Please, somebody have the answer for this...

I am having the same type of errors appearing in /var/log/messages on my Mdk 10.1 AMD64 edition laptop (hp zv5000z)
I noticed it just after building a new kernel (2.6.11) because every ten minutes or so XMMS would hang (my music is on a server on my network). I'll post again if I am able to fix it by using another kernel or tweaking .config.

Here're some of the messages I get:

Mar 15 19:43:29 hyper mount.smbfs[23125]: [2005/03/15 19:43:29, 0] tdb/tdbutil.c:tdb_log(725)
Mar 15 19:43:29 hyper mount.smbfs[23125]:   tdb(/var/cache/samba/gencache.tdb): tdb_lock failed on list 82 ltype=0 (Bad file descriptor)
Mar 15 19:43:29 hyper mount.smbfs[23125]: [2005/03/15 19:43:29, 0] tdb/tdbutil.c:tdb_log(725)
Mar 15 19:43:29 hyper mount.smbfs[23125]:   tdb(/var/cache/samba/gencache.tdb): tdb_lock failed on list 82 ltype=1 (Bad file descriptor)
Mar 15 19:43:29 hyper mount.smbfs[23125]: [2005/03/15 19:43:29, 0] tdb/tdbutil.c:tdb_log(725)
Mar 15 19:43:29 hyper mount.smbfs[23125]:   tdb(/var/cache/samba/gencache.tdb): tdb_lock failed on list 82 ltype=1 (Bad file descriptor)
Mar 15 19:43:59 hyper kernel: smb_add_request: request [ffff81000639ee40, mid=478] timed out!
Mar 15 19:43:59 hyper kernel: smb_add_request: request [ffff81000639ecc0, mid=479] timed out!

Am I right in think this happens if the Windoze machine (that the smbmount is connect to) reboots?

Anyone reading this found the fix to this yet? Surely linux should be able to cope in the same way that any Windoze PC (connecting to a shared drive) would cope?

I was having the same error and restarting my network seemed to fix it. On my Gentoo machine I issued an /etc/init.d/net.et0 restart and my samba was working again.

Well, my error has not gone away despite upgrading to a new distro & kernel (on the client) which makes me think it might be a server side problem. However, other linux and windows users on my network are not experiencing this problem. My computer is AMD64 mobile with Mandiva Linux 2005 LE for x86_64 and the 64bit stuff might be related to the problem. I don't know. I'll watch this thread, but now I've switched to other protocols instead of SMB so I'm not trying to figure this out any more. Good luck all

Originally posted by ksas025 I was having the same error and restarting my network seemed to fix it. On my Gentoo machine I issued an /etc/init.d/net.et0 restart and my samba was working again.

Does it fix it permanently? For me that only fixes it for a few minutes and then it happens again. Also sometimes I have to force a umount of the smbfs and wait a minute before trying again. Just curious...

I had a similar problem, it seems that mounting with cifs instead of smbfs solves it.

mount -t cifs -o ...

instead of

mount -t smbfs -o ...

I was having the same problems with smbfs. Switching to cifs got rid of the error messages, but it still got hung up. I was so frustrated I decided to FTP 2G worth of files over to the linux box I was working on, only to have FTP freeze up randomly as well. I finally thought to check my messages file and discovered it was all due to timeouts on my ethernet device. I popped in a known-good NIC and all is well now. Not sure if this is anyone else's problem, but it's worth checking if you haven't.

The culprit was an onboard sis900 adapter on an ASUS mobo I'm working with. Good thing I didn't pay for it myself or I'd be upset abotu wasting a PCI slot on a (first) NIC.

Good luck to all.

I know this is an old topic, but wanted to bring this up, because this is a bug in kernel 2.6 itself and with every new version released this bug does not get fixed! There are some patches avalabe, but for those who install kernel from .rpm its not the sollution....