[linux-cifs] Re: Problem: mount.cifs has high CPU and "stalls" the using process


  • Previous message:
  • Re: [linux-cifs-client] [PATCH] cifs: hard mount option behaviour implementation
  • Next message:
  • Re: DFS referrals support ?
  • Subject:
  • Re: Problem: mount.cifs has high CPU and "stalls" the using process
  • From:
  • Steve French <smfrench-Re5JQEeQqe8AvxtiuMwx3w <at> public.gmane.org>
  • To:
  • Austin Barton <abarton-kLQwt/ZTpnI <at> public.gmane.org>
  • Cc:
  • "linux-cifs-u79uwXL29TY76Z2rM5mHXA <at> public.gmane.org" <linux-cifs-u79uwXL29TY76Z2rM5mHXA <at> public.gmane.org>, Darren Watts <dwatts-kLQwt/ZTpnI <at> public.gmane.org>, Michael Dillon <mdillon-kLQwt/ZTpnI <at> public.gmane.org>
  • Date:
  • 2015-01-09 19:20:23
Any chance you could get a wireshark or netmon trace (excerpt) of the
high cpu activity (assuming a request is being resent repeatedly
during this interval) - otherwise might be able to clear the kernel
log, enable maximal cifs logging (by echoing 7 into
/proc/fs/cifs/DebugData see -
https://wiki.samba.org/index.php/LinuxCIFS_troubleshooting) and get a
better feel for what is going on in that interval.

Problem is that given how old the kernel is (more than 5 years old) -
it is missing a lot of cleanup and improvements.

On Fri, Jan 9, 2015 at 1:07 PM, Austin Barton  wrote:
> Hello,
>
> I'm looking for some assistance with my mount.cifs issue.
>
> A Windows share is mounted with cifs (rw). At some point a process
> using the mount "stalls". Afterwards mount.cifs is consistently shown
> as the highest item within top having %CPU at approximately 35. The
> process (service) using the mount will not shutdown with init.d and
> must be killed. Outside of the process I'm able to read, list, write
> (w/content), and delete from the mount in question. Please note in the
> "Display Internal CIFS Data Structures for Debugging" below only line
> items 1 and 2 relate to the mount in question.
>
> ------------------------------
> ------------------------------------------------------------------------------------------------------------------
>
> CentOS release 6.5 (Final)
> kernal version: 2.6.32-431.11.2.el6.x86_64
> mount.cifs version: 4.8.1
>
>
> Display Internal CIFS Data Structures for Debugging
> ---------------------------------------------------
> CIFS Version 1.68
> Active VFS Requests: 1
> Servers:
> 1) Name: 10.200.40.132  Domain: I3 Uses: 1 OS: Windows 7 Professional
> 7601 Service Pack 1
>         NOS: Windows 7 Professional 6.1 Capability: 0x1e3fc
>         SMB session status: 1   TCP status: 4
>         Local Users To Server: 1 SecMode: 0x3 Req On Wire: 0
>         Shares:
>         1) \\host.domain\share Mounts: 1 Type: NTFS DevInfo: 0x20
> Attributes: 0xc700ff
> PathComponentMax: 255 Status: 0x1 type: DISK    DISCONNECTED
>
>         MIDs:
>
> 2) Name: 10.200.40.132  Domain: I3 Uses: 1 OS: Windows 7 Professional
> 7601 Service Pack 1
>         NOS: Windows 7 Professional 6.1 Capability: 0x1e3fc
>         SMB session status: 1   TCP status: 3
>         Local Users To Server: 1 SecMode: 0x3 Req On Wire: 0
>         Shares:
>         1) \\host.domain\share Mounts: 1 Type: NTFS DevInfo: 0x20
> Attributes: 0xc700ff
> PathComponentMax: 255 Status: 0x1 type: DISK    DISCONNECTED
>
>         MIDs:
>
>
> 3) Name: 10.200.20.105  Domain: I3 Uses: 1 OS: SpinStream2
>         NOS: Windows 2000 Lan Manager   Capability: 0xd2fc
>         SMB session status: 1   TCP status: 1
>         Local Users To Server: 1 SecMode: 0x3 Req On Wire: 0
>         Shares:
>         1) \\host2.domain\share Mounts: 1 Type: NTFS DevInfo: 0x20
> Attributes: 0x4004f
> PathComponentMax: 255 Status: 0x1 type: DISK
>
>         MIDs:
>
>
> When the stack is inspected it consistently reports the following:
>
> [user@host ~]$ sudo cat /proc/10410/stack
> [] wait_for_response+0xf1/0x130 [cifs]
> [] SendReceive+0x154/0x350 [cifs]
> [] CIFSGetDFSRefer+0x23e/0x670 [cifs]
> [] get_dfs_path+0x65/0x190 [cifs]
> [] expand_dfs_referral+0xef/0x2c0 [cifs]
> [] cifs_mount+0x384/0x690 [cifs]
> [] cifs_get_sb+0xaf/0x2a0 [cifs]
> [] vfs_kern_mount+0x7b/0x1b0
> [] do_kern_mount+0x52/0x130
> [] do_mount+0x2fb/0x930
> [] sys_mount+0x90/0xe0
> [] system_call_fastpath+0x16/0x1b
> [] 0xffffffffffffffff
>
>
>
> Thanks,
> Austin
> --
> To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html



-- 
Thanks,

Steve

This message from: http://www.mailbrowse.com/linux-cifs/10417.html
Previous message: Re: [linux-cifs-client] [PATCH] cifs: hard mount option behaviour implementation
Next message:Re: DFS referrals support ?