Here's the summary for this problem.Many thanks to following people who responded and helped me with this and subsequent problems that arose from this issue. To amerliorate the glm.timeout issue, Matt Hall suggested (and it's a really good one, thanks): ### "When system memory gets low, Solaris unloads unused drivers from memory and reloads drivers as needed ... " So you can try adding forceload: drv/st forceload: drv/sg ### I also duly patched the server and it turns out patch 107460-12, clobbers the st.conf file and fine-tunes how the server and StorEdge communicate (for clarification Solaris 5.7 & 5.8 do not need st.conf mod's only 2.6 does - and there is a yellow insert that came with the L9 that gives you the entries: "QUANTUM DLT8000", "Quantum DLT8000 (in st.conf)", "DLT8k-data"; DLT8k-data = 1,0x38,0,0x10639,4,0x84,0x85,0x88,0x89,3; so what happened to me was after the application of the patch, existing blocks on the tape were not being read effectively (or, were being read a lot more effectively) so the backup's began to fail. I had a mt -f i/o error that was truly bizaare on top of this issue so it made the problem that much harder to diagnose - BaiosThomas at Sun suggested power cycling the StorEdge (duh, I dropped the ball on that one). This on top of expiring the media and reassigning them (ouch!), and a boot -r solved the problem. I'm happily backing up data again. It can only be hoped that st.conf clobbers in patches will not be occurring too frequently, I happen to like having media with data living happily on them (I could have purchased 7 new DLT tapes but didn't). That was fun. Till the next time, David Those who responded (thanks!) Matthew Hall Ivan Fetch Scott Howard Reinaldo Lums Ono Joel Cederholm - Veritas Howard Martin Will Enestvedt Lisa Connell - Sun Baios Thomas - Sun ################################ Original Message: ############################### Howdy, On the Sun web-site, it states quite clearly modifications to the st.conf file must be made when you add-on the Storedge L9 - after some pretty exhaustive searches I cannot find what those modifications should be. I checked the list archives (but am not using a Exabyte, nor have cabling problems etc [BTW - the htdig search implementation is much faster, nicely done]). I'm getting the timeout issue that is most probably due to the lack of these st.conf mod's. Syntax: ######################## Disconnected command timeout for Target 4.0 Apr 15 08:35:47 [serverName] unix: WARNING: ID[SUNWpd.glm.cmd_timeout.6016] ######################### Does any one have rec's for those mods? Will veneratedly summarize. ################################################# ********************************************** David JL Rieger Information Engineering & Strategy University of California, Berkeley PH: 510.643.9666 FX: 510.643.9929 eMail: drieger@socrates.berkeley.edu "The generation of random numbers is too important to be left to chance." Robert Coveyou ********************************************** _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagersReceived on Mon Apr 22 13:30:28 2002
This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:42:41 EST