Search your A-disk for LSEGMAP and PSEGMAP files, they may tell the segment storage is full. VMSES's SEGGEN usage is an area for improvement: all debugging information is missing.
2009/3/17 Shimon Lebowitz <[email protected]>: > On the advice of a respected member of the list, I tried > adding a LOT of virtual storage to MAINT, but even with 512M > I get the same result. This makes sense, since the error message > is not mentioning any problem loading the content of the segment. > > This is happening in VMFBDSEG EXEC at lines 2169-2170. > The first is the SEGMENT RESERVE command, > the second is the invocation of SEGGEN MODULE, which fails. > > Just for fun I commented out the RESERVE, since SEGGEN is apparently > complaining about it, but (no surprise!) SEGGEN did not work then either. > The difference was that there was no message explaining WHY it failed. > > So, here I am stuck, even with 512M, and no segment being built. > Oh, the details of the segment are the defaults, 3M beginning at 20M: > > cp q nss name instseg map > FILE FILENAME FILETYPE MINSIZE BEGPAG ENDPAG TYPE CL #USERS PARMREGS > VMGROUP > 0005 INSTSEG DCSS N/A 01400 016FF SR A 00021 N/A N/A > 0074 INSTSEG DCSS N/A 01400 016FF SR S 00000 N/A N/A > > And catching the SEGMENT RESERVE with an interactive trace, shows: > > 'QUERY SEGMENT *' > Space Name Location Length Loaded Attribute > INSTSEG INSTSEG 01400000 00300000 NO USER > > Any ideas? > > ======== ! ! ! NEWS FLASH ! ! !========= > > For some reason I happened to look at CMSINST LSEGMAP D on the 51D disk, > and *IT* had the answer to the problem!!! One of the files I had listed in > the > build list for the segment was a mistake, and did not exist! > > So, instead of nicely putting out a RC=28 and saying the file was missing, > it hid this message in a file NOT on my A-disk, and expected me to guess the > problem. :-( > Anyway, I made that correction, and then just for fun I reset storage to 64M > before the build. It worked fine! > > Oh... the message about the segment already being reserved is a red herring, > I got that message on the successful run too! > > VMFBLD2760I VMFBLD processing started > VMFBLD1851I Reading build lists > VMFBLD2182I Identifying new build requirements > VMFBLD2182I No new build requirements identified > VMFBLD1851I (1 of 1) VMFBDSEG processing SEGBLIST EXC00000 D, target is > BUILD > 51D (D) > VMFBDS2115I Validating segment INSTSEG > VMFBDS2002I A DEFSEG command will be issued for 1 segment(s). > VMFBDS2219I Processing object INSTSEG.SEGMENT > DMSDCS358E Skeleton segment INSTSEG has already been reserved > HCPNSS440I Saved segment INSTSEG was successfully saved in fileid 0076. > VMFBLD1851I (1 of 1) VMFBDSEG completed with return code 0 > VMFBLD2180I There are 0 build requirements remaining > VMFBLD2760I VMFBLD processing completed successfully > Ready; T=3.73/4.02 12:17:07 > > So I sent all this to the list in case someone else also gets these > messages: > DMSGEN1279E Error(s) occurred during SEGGEN processing. > VMFBDS1965E The command, SEGGEN INSTSEG PSEG A SYSTEM SEGID D2 ( TYPE, > failed > with return code 32 > > Thank you all kindly for your attention :-) > Shimon > -- Kris Buelens, IBM Belgium, VM customer support
