Topic: no domain server to logon v5.6

[Ray Mitchell]

Dear All
I have a working v5.6 server acting as domain controller which has been running OK for a few weeks.One moring Windows workstations could not log on, received message no domain server available. I tried logging on to the console but no response from the server, which then required a forced reboot. After restarting the server worked correctly and users culd log on normally. The system had been working OK about 11.30 pm the previous evening. The only apparently unusual entry in /var/log/messages is the following.

Can anyone tell me what it means and if it is related to the server "malfunction" ??

Thanks
Ray Mitchell

Mar  3 03:31:17 server1 kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000004
Mar  3 03:31:17 server1 kernel:  printing eip:
Mar  3 03:31:17 server1 kernel: c01c938c
Mar  3 03:31:17 server1 kernel: *pde = 00000000
Mar  3 03:31:17 server1 kernel: Oops: 0000
Mar  3 03:31:17 server1 kernel: ppp_deflate zlib_deflate ppp_async printer usbcore appletalk slip ppp_generic
Mar  3 03:31:17 server1 kernel: CPU:    0
Mar  3 03:31:17 server1 kernel: EIP:    0010:[datagram_poll+188/208]    Tainted: P
Mar  3 03:31:17 server1 kernel: EIP:    0010:[]    Tainted: P
Mar  3 03:31:17 server1 kernel: EFLAGS: 00010246
Mar  3 03:31:17 server1 kernel:
Mar  3 03:31:17 server1 kernel: EIP is at datagram_poll [kernel] 0xbc (2.4.18-5)
Mar  3 03:31:17 server1 kernel: eax: 00000000   ebx: c0dbf8e0   ecx: 00000000   edx: 00000000
Mar  3 03:31:17 server1 kernel: esi: 00000049   edi: 00000000   ebp: 00000145   esp: c1927f04
Mar  3 03:31:17 server1 kernel: ds: 0018   es: 0018   ss: 0018
Mar  3 03:31:17 server1 kernel: Process atalkd (pid: 2026, stackpage=c1927000)
Mar  3 03:31:17 server1 kernel: Stack: 0000000a c011f854 c1927f30 c11f0a80 00000000 c2fbb1c0 c01c3e1f c2fbb1c0
Mar  3 03:31:17 server1 kernel:        c24c57a4 00000000 c0146a4c c2fbb1c0 00000000 c1927f54 00000100 c1926000
Mar  3 03:31:17 server1 kernel:        7fffffff 00000008 00000000 0000000a 00000000 c1dd0000 00000001 bffffc14
Mar  3 03:31:17 server1 kernel: Call Trace: [schedule_timeout+20/160] schedule_timeout [kernel] 0x14
Mar  3 03:31:17 server1 kernel: Call Trace: [] schedule_timeout [kernel] 0x14
Mar  3 03:31:17 server1 kernel: [sock_poll+31/48] sock_poll [kernel] 0x1f
Mar  3 03:31:17 server1 kernel: [] sock_poll [kernel] 0x1f
Mar  3 03:31:17 server1 kernel: [do_select+268/576] do_select [kernel] 0x10c
Mar  3 03:31:17 server1 kernel: [] do_select [kernel] 0x10c
Mar  3 03:31:17 server1 kernel: [sys_select+825/1168] sys_select [kernel] 0x339
Mar  3 03:31:17 server1 kernel: [] sys_select [kernel] 0x339
Mar  3 03:31:17 server1 kernel: [system_call+51/64] system_call [kernel] 0x33
Mar  3 03:31:17 server1 kernel: [] system_call [kernel] 0x33
Mar  3 03:31:17 server1 kernel:
Mar  3 03:31:17 server1 kernel:
Mar  3 03:31:17 server1 kernel: Code: 0f ab 48 04 89 f0 83 c4 10 5b 5e c3 90 8d b4 26 00 00 00 00

[Ray Mitchell]

For the benefit of anyone reading this thread.

After a bit of dialogue with Mitel (Thanks very much to them), where they suggested the problem was failing hardware, I initially added more RAM but still had the problem.

I then removed the original RAM and now only have the additional RAM (64Mb) and now I have no more problem.
So most probably the original RAM was faulty.

I notice that Charlie Brady did "give" us a package to test (stress out) our RAM, so my guess is that v5.6 is more demanding on RAM, and if there is a faulty bit somewhere then it will most likely cause a problem.

Hope this helps someone.

Regards
Ray Mitchell

[guestHH]

Hi ray,

I'll give the memtest a try on some other servers. I had it running on a new install as a 'burn in test' for 24 hours without problems.

Thanks fro the pointer.