Re: TSM 4.2.1.6 AIX 4.3.3.0
We have been seeing the server hangs, or something very like them, - as described by Jim (4140) and Henk (4216) - since July at 4140. We have had a PMR open since then and have not made any progress towards a resolution. The latest suggestion I have had is to upgrade to either 4144 or (wait for it) 4212 to fix the problem. This latest suggestion was this week, and after all the problems have been aired with 421x on this list. The finger is being pointed at 'library sharing' code (which we do not use and have library sharing option set to no). I would urge anyone experiencing server problems that require a restart to resolve, to report the problems to IBM/Tivoli. They may not get the focus required to diagnose and fix otherwise. Thanks to all for the information about 421x. I opened a PMR about the licensing problems still persisting at 4216 and was told today that it will be fixed at 4127, which is being built, and that it had been accidentally omitted from 4216. Regards, Sheelagh -- Sheelagh Treweek Oxford University Computing Services Email: [EMAIL PROTECTED] Phone: +44 (0)1865 273205 Fax:-273275 X-Accept-Language: en MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Date: Wed, 31 Oct 2001 09:44:24 -0500 From: Jim Kirkman [EMAIL PROTECTED] Subject: Re: TSM 4.2.1.6 AIX 4.3.3.0 To: [EMAIL PROTECTED] We have seen this exact same scenario with 4.1.4 on OS390 2.10. Backups, q proc, and expiration working, DB backup and dr move cmds invoked but never ran, admin sessions 'frozen' and processes cancelling but never going away. It's only happened once, but then we've only been at 4.1.4 for about a month. I ended up 'bouncing' the server after expiration ended. FWIW Henk ten Have wrote: I wrote: If any bad comes out of this, just let me know too. Well, it didn't take 24 hours. This morning my server was frozen. Simple queries like q mo, q vol, q libvol or q dr never give any result. q se (which worked) showed a lot of frozen admin session, wait time zero, but never ending. q proc (which also worked) showed backup stg processes which didn't do anything for a long time (probably finished, but never ending). The funny thing was that backup's still worked (backup's to diskpools). So finally I had to halt the server and did a restart. The activitylog and the dsmserv.err didn't show anything weird. Cheers, Henk (still not amused) -- Jim Kirkman AIS - Systems UNC-Chapel Hill 966-5884
Re: TSM 4.2.1.6 AIX 4.3.3.0
So, once again, the actual build of the server is not in sync, as I suspected.. This has happened before, and it's very frustrating. I have seen notes on the list from Tivoli-people that they vould look into their routines, and make sure it doesn't happen again. So why do we still have to go through this? On Thu, 1 Nov 2001, Sheelagh Treweek wrote: . . I would urge anyone experiencing server problems that require a restart to resolve, to report the problems to IBM/Tivoli. They may not get the focus required to diagnose and fix otherwise. Thanks to all for the information about 421x. I opened a PMR about the licensing problems still persisting at 4216 and was told today that it will be fixed at 4127, which is being built, and that it had been accidentally omitted from 4216. Regards, Sheelagh -- Sheelagh Treweek Oxford University Computing Services Email: [EMAIL PROTECTED] Phone: +44 (0)1865 273205 Fax:-273275
TSM 4.2.1.6 AIX 4.3.3.0
I wrote: If any bad comes out of this, just let me know too. Well, it didn't take 24 hours. This morning my server was frozen. Simple queries like q mo, q vol, q libvol or q dr never give any result. q se (which worked) showed a lot of frozen admin session, wait time zero, but never ending. q proc (which also worked) showed backup stg processes which didn't do anything for a long time (probably finished, but never ending). The funny thing was that backup's still worked (backup's to diskpools). So finally I had to halt the server and did a restart. The activitylog and the dsmserv.err didn't show anything weird. Cheers, Henk (still not amused)
Re: TSM 4.2.1.6 AIX 4.3.3.0
On 31-Oct-01 Jim Kirkman wrote: We have seen this exact same scenario with 4.1.4 on OS390 2.10. Backups, q proc, and expiration working, DB backup and dr move cmds invoked but never ran, admin sessions 'frozen' and processes cancelling but never going away. It's only happened once, but then we've only been at 4.1.4 for about a month. I ended up 'bouncing' the server after expiration ended. I saw this same scenario with 4.1.X on AIX. And not once, but nearly 4 times a week. That was the (biggest) reason for upgrading to 4.2.1.X. Cheers, Henk.
Re: TSM 4.2.1.6 AIX 4.3.3.0
Henk, I did not have this problem this morning but I can think of a similar situation I'm in with q dr. There are times nothing comes back, the system just sits there. Yesterday I issued a command to do a db backup with RMT1 online. Since nothing was going on it tried to use that drive I'm sure. Once the command was issued the command line did not come back. Yet I queried the activity log and the command was logged. Still the backup did not happen because it tried to use RMT1. Last week I had a specific volser that did not dismount. Since then RMT1, which it was in, has not worked with TSM. I noticed TSM thinks that volser was not in the library so I had to check it in even though it was never removed. TSM has to be restarted each time RMT1 gets a process to do something because I can't cancel the process, well I can but it still sits there doing nothing. The only way to get out of it is to halt TSM and restart. The real strange thing is, since the 25th, I see an entry in the activity log saying the dismount of that same volume has failed. Yet there is no entry in the log that it tried to mount it in RMT1, except for the 25th, the day it did not dismount. I've removed the drive and re-installed it, still the same problem. THis volume has sucesfully been mounted and dismounted in a different drive since this all happened. I ran traces this morning and sent them off to TSM support. I also sent them off to the IBM folks here that work on the hardware, maybe they can help figure out what's going on. For the trace I took all the drives offline and brought RMT1 online. I tried to do an audit vol on a volser. Nothing was mounted, but I happened to notice in the act log TSM tried to mount a total of 3 different tapes in that drive for the audit process. I would guess the reason for this would be file overlap on other volsers, otherwise I'm clueless. Cheers, Henk (still not amused) Me either.. Geoff Gill TSM Administrator NT Systems Support Engineer SAIC E-Mail: [EMAIL PROTECTED] Phone: (858) 826-4062 Pager: (888) 997-9614
