New linux nodes not backing up on schedule but work manually

[mosiac]

I've got an odd issue occuring with any new linux node I attach to my TSM server. I'm currently running 7.1.5 and I have client versions of from 6.2 all the way to 7.1.x.

Many of my old clients are backing up without issue. My newer clients though always seem to miss their schedule and I don't have anything reported in dsmerror.log or dsmsched.log that shows any sort of error or failed connection.

I've made sure the dsmcad service is running, and manual backups using dsmc and running incrementals works fine.

I've made sure the firewall ports are open or the local firewall is disabled for testing but it feels like I'm missing something.

 

[rgg]

Polling or prompted mode? Have you tried switching the mode? If the server prompts the client, then you should see a failure in the actlog of some type if it can't get through. You might also verify the server actually knows the node has a schedule that needs to execute...the undocumented SHOW PENDING command on the server might help you do this.

 

[mosiac]

I've never set modes before I believe it's still the default of any and there is nothing specified in the options file on the node stating a different polling mode. The server does know the node has a schedule by using the show pending command and checking in operations center for it's schedule association.

 

[rgg]

There is no such thing as schedmode of "any". It's either polling or prompted - polling is default and what you seem to be using, which means it reaches out to the server to check for schedules, and then runs them at an agreed upon time.

https://www.ibm.com/support/knowledgecenter/SSGSG7_7.1.6/client/r_opt_schedmode.html

Given you're using polling mode, there has to be something in the dsmsched.log. You should take another look.

 

[mosiac]

I think I found my first problem and that is that the dsmsched.log file stopped recording on may 2nd. I'll get that sorted first and see what shows up once logging works again.

Thanks,