This is a discussion on Re: [Samba] hanging smbd(s) revisited within the Samba forums, part of the Networking and Network Related category; Matt, > To reiterate... the fcntl64 spins in a blocking wait, so we never see an > error unfortunately. spins ...
|
|||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
|
|||
|
Matt,
> To reiterate... the fcntl64 spins in a blocking wait, so we never see an > error unfortunately. spins in a blocking wait? What does that mean? Either it is spinning (chewing cpu) or its in a blocking wait (and not chewing cpu). If its doing both then its a kernel bug. If this is Linux, and its blocked waiting for a lock, then I'd suggest catting /proc/locks while its stuck. From that you should be able to work out the state of each smbd using that tdb, and see if its a "possible" state or not ('possible' defined by the pattern of locks tdb does). Cheers, Tridge -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba |
|
|||
|
Since we moved our data from the NFS server to the Samba server
locally, the freezes have been less. We also get a new message in dmesg which I believe could be related to NFS as well? lease broken - owner pid = 13673 After reading, I assume this means kernel oplocks aren't working? Could this have been our problem while we were resharing NFS as well? The way I understand kernel oplocks, they allow the linux kernel to be aware of samba oplocks for non-samba processes. So, can we disable kernel oplocking and run rsync etc.. while users are working? In which case, is there even a point in having them if the server is strictly used for servering window's clients? kernel: 2.6.9-22 & samba 3.0.21B Thank you, Ryan |
|
|||
|
I don't think anyone is reading this thread any longer but I figured I
would put my progress in it anyway... We upgraded last night to samba 3.0.21c and so far (knock on wood) we haven't had any more error messages about locks or leases. I plan to give it the weekend to make sure and then put our data back on NFS server and see if there is a difference there too. --ryan |
|
|||
|
Ok, upgrading samba didn't work. If anything our problems are worse,
so we are not moving to NFS. I can't figure out if this is locking problems because of new kernel/hardware or microstation special locking needs or what.. I'll post my solution if I ever find one... |
![]() |
| Thread Tools | |
| Display Modes | |
|
|