TSM 7.1.1 Operation Center showing TSM server as unavailable

[Newcomer]

Hi!

I'm very new to TSM and still haven't received any training in our environment since our main technician got sick and wont be back until after Christmas. But none the less i'm the only one except for him that has access at the moment. So the conditions aren't really the best. But anyway, on to the problem;

So we are running TSM 7.1.1 on two linux servers (redhat 2.6.32-504.el6.x86_64) and today when i get to work i see that operation center is showing one of our TSM servers as unavailable. So the first thing i do is to check is the server is down, which it isn't. It's up and running and has some sessions running. But OP-center is still showing it as unavailable.

The one thing that might re related that i can think of is that we had some issues yesterday and had to reboot one of the physical linux servers. But the strage part is that it was the other one, the one that is showing in op-center and working fine.

If i hoover over the server in op-center i get a message saying "the server is running an earlier software version than the hub server. Some functions might be limited". But there haven't been any upgrades as far as i know.
I found some information http://www-01.ibm.com/support/docview.wss?uid=swg21673690#incorr_srv_ver

But the waiting and ctrl + r part didn't really do anyting at all Any tips or ideas on where to start to look for the solution would be really appreciated. But as i said, i'm unfortunately very new.

 

[marclant]

Using a command line, use the QUERY STATUS command on your various TSM Servers to check the version, see if any are back level.

 

[moon-buddy]

Try rediscovering that instance by deleting it from the OP Center and redefining it.

 

[Newcomer]

Thanks for the help.

I tried to remove the spoke server from the monitored group with the command "DELETE GRPMEMBER group_name member_name" which removed it from OP-center as well as from the group when i checked "q monitorsettings". I've added it again from the op-center GUI and it took some time but now it's showing as "normal status". But displaying no capacity for both database, active log and archive log. Which makes no sense to me. I still have the small icon that i can hoover over saying that the software is an earlier version than the hub-server.

 

[moon-buddy]

Thanks for the help.

I tried to remove the spoke server from the monitored group with the command "DELETE GRPMEMBER group_name member_name" which removed it from OP-center as well as from the group when i checked "q monitorsettings". I've added it again from the op-center GUI and it took some time but now it's showing as "normal status". But displaying no capacity for both database, active log and archive log. Which makes no sense to me. I still have the small icon that i can hoover over saying that the software is an earlier version than the hub-server.

Is it than an earlier version?

You can check by running 'q system' on the TSM server in question to see the exact version.

 

[Newcomer]

Both TSM-servers are running "Version 7, release 1, level 1.0" but one thing that i noticed when i connected to the TSM servers were that the time was off by 1-2 minutes according to to the TSM server. But when i logged in directly to the linux-server and checked the date, it was off by around 3-4 minutes. Strange part is that the NTP settings there seems to be correct. Not sure if the timesync errors are related to the op-center issue, but at least i found another error to work on.

 

[Newcomer]

So today i've gotten a new error. It feels like there was some configuration missing.

The problem is that our storage pool has 1962 full volumes and 4 "filling" volumes left. Apparently not a single empty one. The problem is that there is some volumes that is possible to reclaim from. But when i tried with "move data volumename reconstruct=yes" it failed since we dont have a free volume to move to. I think the issue here is that we have replication active from one of our other TSM servers.

So the question is if its possible to delete the volumes on the targeted replication server which contains the replicated data from our other TSM server?
If i somehow can find a volume that contains only replicated data that already exists on the "source" server, and delete it so that i can start some reclamation process and free up some volumes.

I've already stopped the replication for now, since i think we need to tune it and not replicate "everything" as we do now.

 

[Newcomer]

*Update*

I've had quite a few errors during the Christmas time. But now i'm starting to get back on track. But i think that the problem is that the server that is showing as "unavailable" has the wrong time. The server is running on Red Hat 6.6 and the time is currently 2-3 minutes before correct time.

What i'm wondering is if i need to halt the TSM-server before i correct the time on the server?
My guess would be that it's best to halt the TSM-server since it's such a large time difference, but does anyone have actual information on the "best practice"?