Media Write Errors

[jdblock]

We are having a problem with media. We are starting to get a lot of media write errors. I believe it is media rather than the drive because when we freeze the tape that is being used when the error occurs, and re-run the backup, it finishes successfully. The real problem that we are having is when a backup gets this error on the first try, it uses the same tape for the second try and fails again. We would like to force it to use another tape on the retry. Can we automatically freeze the tape on an error 84 in NetBackup? What other methods can we use to force it to another tape on the retry?

Setup:
NetBackup DataCenter 4.5
Master Server and Media Servers are UNIX (Solaris)
Mix of UNIX and Windows Clients

Thanks,
John Block

 

[PGPhantom]

Create the following files under the veitas\netbackup directory:

DRIVE_ERROR_THRESHOLD - Simply enter in a numer on its own e.g. 2
MEDIA_ERROR_THRESHOLD - Same as the drive error.
TIME_WINDOW - Enter in a number to specify hours. e.g. If you enter in 8 and the MEDIA_ERROR_THRESHOLD is 2 what will happen is that if you get 2 media errors in 10 hours, the tape will be frozen.

One issue that I have found ... A bug, If the tape is loaded into drive #1 and gets an error and is then loaded into drive #2 and gets an error, the error count is 1 and not 2 as it should be. If the same tape is loaded into the same drive - Bingo = Frozen

Case sensitivity is paramount - Be sure that the files are all upper case.

Let me know if this works for you.

 

[adpwl1]

Also, what is the tape format you are using? DLT/LTO 1 or 2. If LTO, IBM/Seagate/HP

There is a problem with IBM LTO-2 drives that is corrected in a new firmware.