system reboot every night

[Umesh Patel_3]

We have ES40 system, with tru64 5.1a and firmware 5.57.

The problem is the sytem keep reboot. I am not sure the reason.

here is the info from crashdata info
#
# Crash Data Collection (Version 1.4)
#
_crash_data_collection_time: Tue May 25 14:43:23 PDT 2004
_current_directory: /
_crash_kernel: /var/adm/crash/vmunix.16
_crash_core: /var/adm/crash/vmzcore.16
_crash_arch: alpha
_crash_os: Compaq Tru64 UNIX
_host_version: V5.1A (Rev. 1885)
Compaq Tru64 UNIX V5.1A (Rev. 1885); Wed Aug 21 15:05:58 PDT 2002
_crash_version: V5.1A (Rev. 1885)
Compaq Tru64 UNIX V5.1A (Rev. 1885); Wed Aug 21 15:05:58 PDT 2002


warning: Files compiled -g3: parameter values probably wrong
_crashtime: struct {
tv_sec = 1085519544
tv_usec = 6832
}
_boottime: struct {
tv_sec = 1085498713
tv_usec = 696864
}
_config: struct {
sysname = "OSF1"
nodename = "odalpha.us.oracle.com"
release = "V5.1"
version = "1885"
machine = "alpha"
}
_cpu: 57
_system_string: 0xffffffffff91d030 = "Compaq AlphaServer ES40"
_ncpus: 4
_avail_cpus: 4
_partial_dump: 1
_physmem(MBytes): 12287
_panic_string: 0xfffffc00007c4118 = "Processor Machine Check"
_paniccpu: 3


----

Machine Check Processor Fatal Abort
Machine check code = 0x100000098
Ibox Status = 0000000000000000
Dcache Status = 0000000000000008
Cbox Address = 0000000297528400
Fill Syndrome 1 = 00000000000000be
Fill Syndrome 0 = 000000000000002b
Cbox Status = 000000000000001b
EV6 captured status of Bcache mode = 0000000000000009
EV6 Exception Address = 0000000121de8408
EV6 Interrupt Enablement and Current Processor mode = 0000007ee0000008
EV6 Interrupt Summary Register = 0000000080000000
EV6 TBmiss or Fault status = 0000000000000290
EV6 PAL Base Address = 0000000000018000
EV6 Ibox control = fffffe0007304396
EV6 Ibox Process_context = 0000728000000004
O/S Summary flag = 0000000000000004
Cchip Base Address (phys) = 00000f01a0000000
Cchip Device Raw Interrupt Request = 0000000000000000
DRIR Register Decode:
PCI Device Interrupt Mask = 0000000000000000
Cchip Miscellaneous Register = 0000000000000000
Misc Register Decode:
Cchip Revision: 00
ID of CPU performing read: 00
Pchip 0 Base Address (phys) = 00000f0180000000
Pchip 0 Error Register = 0000000000000000
Pchip Error Register Decode:
PCI Xaction Start Address = 0000000000000000
PCI Command: Interrupt Acknowledge
Pchip 1 Base Address (phys) = 00000f0380000000

Pchip 1 Base Address (phys) = 00000f0380000000
Pchip 1 Error Register = 0000000000000000
Pchip Error Register Decode:
PCI Xaction Start Address = 0000000000000000
PCI Command: Interrupt Acknowledge
CPU 3 is prevented from being rebooted.
The system must be reset or power cycled to clear this state.
panic (cpu 3): Processor Machine Check
syncing disks... done


--------

[Michael Schulte zur Sur · Accepted solution]

Hi,

reseat cpu 3 and if you have maintenance contract let HP replace it.

greetings,

Michael

[Umesh Patel_3]

Thanks Mike,

We do not have maintanance contract, so if we take out cpu 3 then will it solve the problem?

[Michael Schulte zur Sur]

Hi,

Machine Check Processor Fatal Abort
CPU 3 is prevented from being rebooted.
panic (cpu 3): Processor Machine Check

as you can see, it is not working correctly, probably has a hardware problem. If you can work without, do it. Chances are, it will work.

greetings,

Michael

[Ravi_8]

HI

CPU 3 is causing the problem, before calling the HP.you can disable the CPU 3 and check. If there is a condition that you have to work with all 4 processors, then log a call with HP else continue with remaining 3 CPU's

you can put the processor 3 off line by:
#psradm -f 3

for more info refer man page of psradm

[Umesh Patel_3]

Thanks!

I have put it offline, let see what happen now! I will post the result.

Thanks
Umesh

[Umesh Patel_3]

Seems like not working.

Now it give the same message with cpu 2.

Do you think the air condition in the room make this problem?

[Ralf Puchner]

open a call within the support center, the cpu seems to be damaged!