EXIT STATUS 74: client timed out waiting for bpstart_notify to complete

[Anth105]

Here is my Env;

NBU 6.5.2A
Master Server running on AIX 5.3
Media Server/Client running on Win2k

Problem: Client backup fails with status code "EXIT STATUS 74: client timed out waiting for bpstart_notify to complete"

Troubleshooting :
Increase the BPSTART_TIMEOUT to 3600 in the bp.conf on the Media Server/Client
Reboot the Media Server/Client to cleaning up any hanging process
Follow the instructions as per TN http://support.veritas.com/docs/300456 and the troubleshooting guide for information on error code 74 on page 181: http://support.veritas.com/docs/290230
Re-run the backup without succes.

I have a opened case with Symantec but the support guy has not come up with nothing new than the troubleshootings steps that I have already carried out.

PS: bpstart_notify script has not been customised.

I wonder if someone has any ideas

Many thanks in advance

Anth105

[Stumpr2]

What do you mean that it has not been customised? What are you trying to do with the script? Try running it from the command line.

[Marianne · Accepted solution]

If you have copied the sample bpstart_notify from the goodies directory on the master to the client's netbackup/bin without modifying it, it will produce status 74.
If you intend to use it, customize it as per your requirements and test from cmd line.
If not, delete it.

[John_Stockard]

Also note that if you increase the BPSTART_TIMEOUT value to 3600 seconds, you also need to increase the CLIENT_READ_TIMEOUT value to also be 3600 seconds.  See Technote 314768 for details.

Also, make sure you put the BPSTART_TIMEOUT and CLIENT_READ_TIMEOUT settings on the NetBackup media server that is trying to process the backup, not on the client.  See Technote 275615.

Some other words of advice regarding bpstart_notify and bpend_notify scripts:

• For best results, make sure the script and the commands issued by the script don't produce any output to the screen (stdout) when you run them.  Obviously, you'll want to fix the script to do this after you're done troubleshooting it from the shell prompt or command line.  It's quite hard to troubleshoot a script that produces no output.  :)

• A best practice for the bpstart_notify and bpend_notify scripts is to hard-code an exit value of 0 so that NetBackup knows that the script completed successfully.  A Unix/Linux client can do this by putting the line "exit 0" (without the quotes) at the end of the script.  A Windows client can do this by putting the line "@echo 0 > %6" (without the quotes, note that there is no space between % and 6 or between the @ and "echo").  The %6 in that line is a parameter that NetBackup passes to the script and is the name of a file that NetBackup checks to see if the script completed successfully.  If the file exists and contains the character "0", NetBackup assumes that the bpstart_notify.bat script completed successfully and it proceeds with the backup.  This is vaguely documented in the NetBackup 6.5 System Administrator's Guide for Windows, Volume 2 on page 73.

[Anth105]

I resolved the problem by copying the bpstart_notify from a working node to the offending machine.

Many thanks to all for your contribution