Data getting corrupted on HP SureStore Fc60 Disk Array

[Naveen Shetty]

Hi All ,

Kindly if someone can suggest me the solution for the below listed problem .

I am having a FC60 Disk Array box with 2 no.of SC10 with it , with 9*73GB RAID5 on each SC10 .

I am finding that data is getting corrupted very frequently . That is , I am finding few rows in the oracle database are getting deleted on its own .

Patch bundles applied are of June 2000 . It was also tried by applying December 2000 Patch Bundles .

Kindly guide me with the solutions ...!

Thanks in advance ...!

Naveen Shetty.

[A. Clay Stephenson]

Hi Naveen,

I think we need a little more data first.

1) Have you checked the system logs for errors?

2) Are you running raw disk? Asynch_IO ? VxFS or hfs?

3) Are your applications crashing? Anything in the alert.ora logs?

4) Are the rows only disappearing from one table?

5) Is only one array showing the problem.

At this point, I can't say whether it's hardware, OS, or application software. We need a bit more data. The above list should get us started so that we can narrow the focus.

Regards, Clay

[Insu Kim]

Data corruption issue has been found on FC60 with controller firmware HP08 or greater.
Do "amdsp -a " to identify which level of controller firmware has been applied for your disk array and if it's HP08 or greater, I would recommend that you place a support call to HP to get help or you can update controller firmware to HP03 on your own to insure data integrity.

Please check that the following patch is in the system which is mass storage patch and if it's not, apply it as soon as possible.

PHKL_21834 ( HPUX 11.0 )

Best regards,

[Naveen Shetty]

Hi ,

I have confirmed that Firmware version on FC60 is HP03 .Is there any such known problems with even this version of firmware .

Also I have checked that PHKL_21834 patch is installed in the server .

Hi Stephenson ,

Kind find the details reqd by you as below :
1. System logs are not showing any errors .
2. I am running VxFS on this .
3. Situation was not allowed to go upto the applications crashing as the server is working with very critical applications .
4. No the rows are not disappearing from only one table , it happens with any oracle tables .
5. I will revert back on the fifth query .

Kindly suggest me ........!

Thanks in advance ..!

Naveen Shetty .

[Alexander M. Ermes]

Hi there.
Excuse this question, but have the data been committed ?
Have any triggers been fired / procedures been started etc ?
Try to put triggers on the tables, which write data to a log, if a record is deleted and the delete is committed or setup a trace in Oracle.
Rgds
Alexander M. Ermes

[Carlos Fernandez Riera]

Are you runnig oracle 8.1.7 plus async_disk?...

A new note on metalink is available on this situation from today or few days ago.

I have experienced that error on my server, and now is a known bug.

Please search metalink or call you Oracle support.

Good Luck.