julie@levell.austin.ibm.com (Julie A. Levell) (03/02/91)
This was posted in comp.unix.aix - care to comment? jerry ------------------------------------------------------------------------- In article <1991Feb28.143052.15422@ux1.cts.eiu.edu> you write: > > Hi all! We recently had some trouble with our rs6000/530 and had to >reinstall the thing from scratch (the original floppies...all 78 of 'em). >I also restored some stuff from other backup tapes including the /usr >subdirectory, the /etc subdirectory and the /u subdirectory; to catch up >with where we were before our problems. > Anyway, I've still got one problem with the volume group. When I boot >I get measseges saying : lvaryon: the volume group cannot be varied on >because there are no good copies if the descripter area. This message >results when I do a varyonvg rootvg. Other messages I get while in smit: > getlvodm: Unable to find device if <numbers> in the Device Configuration >Database. Have you, by chance, seen something like this? The last >thing "Software Support" had to say to me was to pull the battry off the >NVram for at least 45 min. and let it rebuld itself. Any thoughts? >Just what is going to happen if I unplug that battery? Is there any other >way to fix this? This problem doesn't seem to interfere with the >functioning of the machine but I don't get any info on the volume group >(lsvg rootvg - tells me to varyon the VG and shows ????? everywhere else.) > I talked to "Software Support about my problem...It seems >that software support is really "Software Defects Support". Once they >decided my problems were not "defects" they told me to call my local >SE and/or CE. Unfortunatly my SE is really a marketing rep and is little >help when it comes to problems. Thanks for any suggestions! > >TTFN > >-- > Dave Beedle Office of Academic Computing > Illinois State University > Internet: dbeedle@rs6000.cmp.ilstu.edu 136A Julian Hall > Bitnet: dbeedle@ilstu Normal, Il 61761 Well, it is a database problem. You've copied your old cfg database over your new one, which is usually fine for everything but lvm, because when you re-installed and created a new rootvg, it got a new volume group identified. So the easiest way to fix is (not really easy): 1. Use odme to remove all the old lvm info. Remove rootvg, hd1, hd2, hd3, hd4, hd5, hd6, hd7, hd8, and any lv's you've created in the rootvg. odme CuDv odme CuAt odme CuDep odme CuDvDr Use PF5 to delete each one. You'll know then when you see them. 2. Then run these two commands. redefinevg -d hdisk0 rootvg varyonvg -m1 rootvg And that will do it. Sorry there isn't an easier way, but we didn't anticipate this in our testing. Call if you have any questions. Julie -- *=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*= Julie A. Levell IBM Advanced Workstation Division Austin, Texas IBM Internet: julie@levell.austin.ibm.com IBM VMNET: JULIEL at AUSVMQ DeskNet: 4C-29/994 SpeakNet: 823-5178 (Tie 793-5178)