[DRBD-user] sock_sendmsg time expired

Steven Wilton steven.wilton at team.eftel.com
Tue Nov 16 08:57:43 CET 2004


Hi,

I've just set up drbd on our systems, and have hit the following 
problem.  After setting up the drbd config, and running
"drbdsetup /dev/drbd5 primary --do-what-I-say" to make one of the nodes 
primary, the initial sync did not start.  On the primary server, I got 
the following logs from the kernel:
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967046
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967045
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967044
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967043
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967042
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967041
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967040
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967039
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967038
drbd4: [pdflush/1583] sock_sendmsg time expired, ko = 4294967037


On the secondary server, I got the following logs from the kernel:
------------[ cut here ]------------
kernel BUG at kernel/timer.c:266!
invalid operand: 0000 [#2]
Modules linked in: drbd ipt_REJECT iptable_nat ip_conntrack 
iptable_filter ip_tables aic7xxx ips eepro100 e100
CPU:    0
EIP:    0060:[<c0119170>]    Not tainted VLI
EFLAGS: 00010046   (2.6.9)
EIP is at mod_timer+0x10/0x50
eax: e92677dc   ebx: e92677dc   ecx: 00000001   edx: e9267838
esi: 001a0e8a   edi: 00000002   ebp: e92674b0   esp: e2707ec0
ds: 007b   es: 007b   ss: 0068
Process drbdsetup (pid: 3148, threadinfo=e2707000 task=ea9416b0)
Stack: 0000000e e92674b0 f09cf41e e92677dc 001a0e8a 00000282 e8d12e40 
ee58f000
        f09c2f69 e92674b0 00000002 e8d12eac bffffc8c e8d12e40 ecf479c0 
e92675f4
        e84fcb4c 00000000 00000000 0804a878 00000004 c010e9b8 bffffcfc 
ef86472c
Call Trace:
  [<f09cf41e>] _set_cstate+0x9e/0xcc [drbd]
  [<f09c2f69>] drbd_ioctl+0x3e9/0xa80 [drbd]
  [<c010e9b8>] do_page_fault+0x0/0x518
  [<c028d286>] blkdev_ioctl+0x35a/0x36e
  [<c0146cbd>] block_ioctl+0x1d/0x24
  [<c014f099>] sys_ioctl+0x1cd/0x1e4
  [<c01043ed>] error_code+0x2d/0x38
  [<c01039e3>] syscall_call+0x7/0xb
Code: ff ff 83 c4 04 9c 5b fa 56 57 e8 88 fe ff ff 89 7e 1c 83 c4 08 53 
9d 5b 5e 5f c3 56 53 8b 5c 24 0c 8b 74 24 10 83 7b 14 00 75 08 <0f> 0b 
0a 01 49 42 33 c0 81 7b 10 6e ad 87 4b 74 09 53 e8 09 fe
  <3>drbd4: short sent PingAck size=8 sent=-1001
Unable to handle kernel NULL pointer dereference at virtual address 00000004
  printing eip:
c011a882
*pde = 00000000
Oops: 0000 [#3]
Modules linked in: drbd ipt_REJECT iptable_nat ip_conntrack 
iptable_filter ip_tables aic7xxx ips eepro100 e100
CPU:    0
EIP:    0060:[<c011a882>]    Not tainted VLI
EFLAGS: 00010002   (2.6.9)
EIP is at force_sig_info+0x2a/0x68
eax: 00000004   ebx: e9299110   ecx: 00000000   edx: 00000000
esi: 00000283   edi: 00000001   ebp: 00000002   esp: e9296f70
ds: 007b   es: 007b   ss: 0068
Process drbd4_asender (pid: 2597, threadinfo=e9296000 task=ea98cbd0)
Stack: e92674b0 e92678a8 00000000 c011af11 00000001 00000001 e9299110 
f09cf6b9
        00000001 e9299110 00000008 e92674b0 e9267760 f09cb190 e92678a8 
00000001
        00000000 e92678f0 e92674b0 e92678f8 00000000 0000000c 00000008 
00000008
Call Trace:
  [<c011af11>] force_sig+0x11/0x18
  [<f09cf6b9>] _drbd_thread_stop+0x91/0x13c [drbd]
  [<f09cb190>] drbd_asender+0x3c0/0x3ea [drbd]
  [<f09cf4d3>] drbd_thread_setup+0x87/0xa8 [drbd]
  [<f09cf44c>] drbd_thread_setup+0x0/0xa8 [drbd]
  [<c01020a5>] kernel_thread_helper+0x5/0xc
Code: f6 57 56 53 8b 7c 24 10 8b 5c 24 18 9c 5e fa 8d 4f ff 0f a3 8b e4 
04 00 00 19 c0 85 c0 75 12 8b 83 e0 04 00 00 8d 14 89 83 c0 04 <83> 3c 
90 01 75 21 8b 83 e0 04 00 00 8d 14 89 83 c0 04 c7 04 90


  I noticed that there were a few threads related to the sock_sendmsg 
time expired, error, but did not see the second error in the archive.

Is there enough information above to debug the problem?

thanks

Steven



More information about the drbd-user mailing list