My problem nodes are running RedHat 6.2; TSM client v4.2.3; ext2 filesystems.
Server is Windows 2k server, SP2; TSM server 5.1.1
For some reason, the 'dsmc sched' process is not backing up the most important part of some of my machines. This would be the portion containing customer data.
There are no entreis in the errorlog on the node that indicate anything amiss. The schedlog does not contain anything to indicate a complete process. The actlog on the server says the session had a timeout and cancelled the session.
Here's the thing that makes me go 'hmm'. I registered a new node (appropriately enough called 'test') for temporary reasons and started dsmc with the 'virtualnodename=test' option, and manual incrementals work.
This whole paragraph is behavior withouth the 'virtualnodename' flag: The problem node only backs up a few files (it inspects less than 20), which I *KNOW* to be incorrect because there are little more than 1.7million files on that filesystem, and 20 files in an incr seems extremely low since I work for a webhoster and there are apache access and error logs (not excluded from backups) in that filesystem. We've even manually editted files to add content and we see the same bahavior. The manual incr returns in a matter of 10 seconds or so, which again I find to be extremely fast for traversing an active filesystem with over 1.7mil files.
On this machine, the filesystem in question is rather large (129G), but so are the rest of my nodes and they experience no problems. Could this have something to do with my issues? I've checked the inclexcl on both nodes, and they match each other. Actually, they match the other 20-some-odd nodes I have too.
Thanks for taking the time to read this, and thanks in advance for any feedback, assistance or help.
Server is Windows 2k server, SP2; TSM server 5.1.1
For some reason, the 'dsmc sched' process is not backing up the most important part of some of my machines. This would be the portion containing customer data.
There are no entreis in the errorlog on the node that indicate anything amiss. The schedlog does not contain anything to indicate a complete process. The actlog on the server says the session had a timeout and cancelled the session.
Here's the thing that makes me go 'hmm'. I registered a new node (appropriately enough called 'test') for temporary reasons and started dsmc with the 'virtualnodename=test' option, and manual incrementals work.
This whole paragraph is behavior withouth the 'virtualnodename' flag: The problem node only backs up a few files (it inspects less than 20), which I *KNOW* to be incorrect because there are little more than 1.7million files on that filesystem, and 20 files in an incr seems extremely low since I work for a webhoster and there are apache access and error logs (not excluded from backups) in that filesystem. We've even manually editted files to add content and we see the same bahavior. The manual incr returns in a matter of 10 seconds or so, which again I find to be extremely fast for traversing an active filesystem with over 1.7mil files.
On this machine, the filesystem in question is rather large (129G), but so are the rest of my nodes and they experience no problems. Could this have something to do with my issues? I've checked the inclexcl on both nodes, and they match each other. Actually, they match the other 20-some-odd nodes I have too.
Thanks for taking the time to read this, and thanks in advance for any feedback, assistance or help.
