Re: Complete Network Loss

[Carsten Otto]

On Sat, Sep 06, 2008 at 06:50:24PM +0200, Carsten Otto wrote:
> After two minutes without connectivity dmesg shows that some application
> (e.g. rsync) is stalled, but gives no further information.

Like this:

INFO: task vsftpd:10885 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
vsftpd        D c0666e04     0 10885  10883
       f673a740 00200082 f5f80ec0 c0666e04 c0669800 c0669800 f626e080 f626e1d4 
       c2010800 00000000 c040eb31 f626e1d4 c040f243 f63c36c0 f63c36c0 f5cd4464 
       00000000 f575111c c0239f8d f5cd4400 f63c36c0 f54f3e80 c0611e84 c2010800 
Call Trace:
 [<c040eb31>] ip_local_out+0x15/0x17
 [<c040f243>] ip_queue_xmit+0x2b6/0x2f8
 [<c0239f8d>] request_send_nowait_locked+0x74/0x85
 [<c047abcc>] io_schedule+0x21/0x2b
 [<c013f3a2>] sync_page+0x33/0x36
 [<c047ad58>] __wait_on_bit_lock+0x31/0x59
 [<c013f36f>] sync_page+0x0/0x36
 [<c013f35f>] __lock_page+0x77/0x7f
 [<c012eaab>] wake_bit_function+0x0/0x3c
 [<c0175846>] __generic_file_splice_read+0x228/0x39a
 [<c047c314>] _spin_lock_bh+0x8/0x1e
 [<c03f10b8>] lock_sock_nested+0x88/0x90
 [<c047c314>] _spin_lock_bh+0x8/0x1e
 [<c03f0fc0>] release_sock+0xf/0x7f
 [<c0415c8a>] tcp_sendpage+0x51a/0x526
 [<c03ee662>] sock_sendpage+0x31/0x36
 [<c01741af>] pipe_to_sendpage+0x4f/0x59
 [<c01747aa>] spd_release_page+0x0/0xa
 [<c0175a2d>] generic_file_splice_read+0x75/0x8f
 [<c01745a0>] do_splice_to+0x4a/0x58
 [<c017495a>] splice_direct_to_actor+0xcd/0x19f
 [<c017453e>] direct_splice_actor+0x0/0x18
 [<c0174a9f>] do_splice_direct+0x73/0x93
 [<c015b701>] do_sendfile+0x187/0x1e5
 [<c015b844>] sys_sendfile+0x69/0x7c
 [<c0102bee>] syscall_call+0x7/0xb
 =======================
-- 
Carsten Otto
carsten@xxxxxxxxx
www.c-otto.de

Attachment: signature.asc
Description: Digital signature