continous error message in syslog

[laiju.c.babu]

Hi Team,

In my system's syslog i am getting continous error message. Please find below the error message

Sep 11 23:37:30 ***** vmunix: function == 0xE00000000054E4E0, arg == 0xE00000013A02A800, ticks == 0xFE8B24F1,  flags == 0x0
Sep 11 23:37:30 ***** vmunix: function == 0xE00000000054E4E0, arg == 0xE00000013A02A800, ticks == 0xFE8B24F0,  flags == 0x0

     [14]> uname -a
HP-UX ***** B.11.23 U ia64 4021640879 unlimited-user license

The above error message is repeating continously .

Thanks a lot for the  help

Regards

[Patrick Wallek · Accepted solution]

This is a known problem with the SCSI Ultra320 driver once the system has been up for 400+ days or so.

See this thread:

http://h30499.www3.hp.com/t5/System-Administration/syslog-log-filling-with-error-messages/td-p/4662395#.UE_7VIqPWp0

The latest ultra320 driver can downloaded from:

https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=scsiU320-00

[laiju.c.babu]

Hi Patrick,

Thanks a lot for the  information.

From the cstm i came to know in my server i have two MPT SCSI adapaters

 25  0/1/2/0              MPT SCSI Adapter (HP_A717 Information Successful
 27  0/1/2/1              MPT SCSI Adapter (HP_A717 Information Successful

And i got the below information for these two adapters.

-- Information Tool Log for MPT SCSI Adapter on path 0/1/2/0 --

Log creation time: Wed Sep 12 12:13:42 2012

Hardware path: 0/1/2/0

Device File               : /dev/mpt8
Adapter                   : PCI-X DUAL CHANNEL ULTRA320 SCSI ADAPTER
Specification             : PW=15W PCI-X=64BIT,133MHZ
Product ID                : A6961-60111
Device ID                 : 0x0030
Vendor ID                 : 0x1000
PCI Revision ID           : 8
Subsystem ID              : 0x12c5
Subsystem Vendor ID       : 0x103c
Firmware Version          : 01.03.35.70 IO
Firmware Image Size       : 40632 bytes
SCSI Initiator ID         : 7
PCI Configuration Space Registers
   Status = 0x0230  Command = 0x0157
   Class Code = 0x10000 (SCSI Controller)
   Cache Line Size = 0x20
   Latency Timer = 0x40
   Expansion ROM Address = 0x90100000

-- Information Tool Log for MPT SCSI Adapter on path 0/1/2/1 --

Log creation time: Wed Sep 12 12:13:42 2012

Hardware path: 0/1/2/1

Device File               : /dev/mpt9
Adapter                   : PCI-X DUAL CHANNEL ULTRA320 SCSI ADAPTER
Specification             : PW=15W PCI-X=64BIT,133MHZ
Product ID                : A6961-60111
Device ID                 : 0x0030
Vendor ID                 : 0x1000
PCI Revision ID           : 8
Subsystem ID              : 0x12c5
Subsystem Vendor ID       : 0x103c
Firmware Version          : 01.03.35.70 IO
Firmware Image Size       : 40632 bytes
SCSI Initiator ID         : 7
PCI Configuration Space Registers
   Status = 0x0230  Command = 0x0157
   Class Code = 0x10000 (SCSI Controller)
   Cache Line Size = 0x20
   Latency Timer = 0x40
   Expansion ROM Address = 0x90000000

From the above information Firmware version is specified as "Firmware Version          : 01.03.35.70 IO". Is this is the firmware version which you are pointing to me .

Regards

[Team_of_5]

Hi Laiju,

This issue we have seen on servers which all are  up for more days [approx. above 400 days].

Your syslog size will increase very fastly due to this error as well !!!! be cautious!!!!

Permant FIx you have to apply the SCSI patch which Mr. Patric has suggested.

Workaround : Take a reboot of the machine, this will also address the issue. 

Regards,

MC

[Laurent Menase]

You should also check

echo  0xE00000000054E4E0/a | adb -o /stand/vmunix /dev/kmem

if the result is not

mpt_os_callback:

mpt_os_callback:mpt_os_callback:

Then it is not the right cause.

If the result is str_timeout,  it is not that error.

Known problem which can cause the same:

VERITAS Incident Number: 1413790   /Veritas Storage Foundation for Oracle RAC 4.1

http://www.symantec.com/business/support/index?page=content&id=TECH67726

VERITAS Incident Number: 14139223   /Veritas Storage Foundation for Oracle RAC 5.1 

http://sfdoccentral.symantec.com/sf/5.1SP1/hpux/html/sfrac_notes/ch01s09s01.htm

If the   address do not match either with mpt_os_callback nor str_timeout,
If the address matches with str_timeout and you don'"t have VSF installed,

then you should contact hp support,

Anyway the workaround of rebooting every 400days works for all those known possible causes

[laiju.c.babu]

Hi Laurrent,

Thanks a lot for the information you provided. This time i just selected the work around. I  rebooted the server and now the error is not appearing.

I will follow your steps if similar issue happens for any othe server.

Thanks a lot ...

Regards