Re: HELP: Unknown Error Messages From oncheck

In article <E4xoy1.CLq@harriet.rencorp.com>, Yagnesh <ren1999@harriet.rencorp.com> writes > >On running oncheck -cI on some of our big tables (~8-10GB) the following >error messages were reported by oncheck. All the indexes that produced such >error message were dropped and recreated. On running oncheck -cI on new >indexes, same error messages were produced. > >Any help as to how to fix this will be appreciated. > >Following are the details..... > > >Here's the output from oncheck that was done after the indexes were rebuilt. > >-------------------------------------------------------- >Fri Jan 31 12:59:52 MST 1997 > >Validating index pcla_ix_01 for rims:prodadmn.ptchg_ldgr_a... > Index pcla_ix_01 >Could not check rowids and perform data<->index check >Unknown error message 0. Sounds like either a) a bug in oncheck b) something more fundamental is wrong then just index/table corruption e.g. reserved pages are corrupt, extent information is corrupt orsystem catalogs are corrupt Try oncheck -cr oncheck -ce oncheck -cc <database name> >Fri Jan 31 14:23:22 MST 1997 > >Validating index pcla_ix_02 for rims:prodadmn.ptchg_ldgr_a... > Index pcla_ix_02 >Could not check rowids and perform data<->index check >Unknown error message 0. >Fri Jan 31 15:24:50 MST 1997 >-------------------------------------------------------- > >At the same time these messages appeared in the online.log file... > >-------------------------------------------------------- >14:23:22 I/O bad request chunk 3156, pagenum 228477, pagecnt 1 >14:23:22 I/O bad request chunk 3156, pagenum 228477, pagecnt 1 Again either extent/reserved info is corrupt so online thinks part of extent is off the end of a chunk or you have hit a bug. >14:28:22 Checkpoint Completed: duration was 0 seconds. >14:33:23 Checkpoint Completed: duration was 0 seconds. >14:38:23 Checkpoint Completed: duration was 0 seconds. >14:43:26 Checkpoint Completed: duration was 0 seconds. >14:48:27 Checkpoint Completed: duration was 0 seconds. >14:58:28 Checkpoint Completed: duration was 301 seconds. >15:03:30 Checkpoint Completed: duration was 0 seconds. >15:08:31 Checkpoint Completed: duration was 0 seconds. >15:13:31 Checkpoint Completed: duration was 0 seconds. >15:18:33 Checkpoint Completed: duration was 0 seconds. >15:24:49 Checkpoint Completed: duration was 77 seconds. >15:24:50 I/O bad request chunk 3157, pagenum 64637, pagecnt 1 >15:24:50 I/O bad request chunk 3157, pagenum 64637, pagecnt 1 >--------------------------------------------------------> >Before rebuilding the indexes, the oncheck had produced same error messages >for all the indexes but one. One index produced assret failed message in the >log that and here it is.... > >-------------------------------------------------------- >11:41:53 bfcheck: bad page: pg_flags 4 != type 10>buffer header >c0cf9fc0: c0cfa048 c0cf9f38 00000000 00000000 ...H...8 ........ >c0cf9fd0: 00000001 00000000 00000000 00000000 ........ ........ >c0cf9fe0 * >c0cf9ff0: 00000000 00000000 00000000 00000000 ........ ........ >c0cfa000: 0001638e 07d10001 00006266 5b323534 ..c..... ..bf[254 >c0cfa010: 33395d00 00000000 00000000 c0b63a50 39]..... ......:P >c0cfa020: c07a8a18 c0024818 c0be5bc0 00020000 .z....H. ..[..... >c0cfa030: 001e0000 005cad54 c3fb0000 c5da9578 .....\\.T .......x >c0cfa040: 00000000 00000080 ........ >page >c3fb0000: 005cad54 7676db8a 00000004 001807e4 .\\.Tvv.. ........ >c3fb0010: 00000000 00000000 ffffffff fffffffe ........ ........ >c3fb0020: ffffffff ffffffff ffffffff ffffffff ........ ........ >c3fb0030 * >c3fb07f0: ffffffff ffffffff 00000000 7676db8a ........ ....vv.. >11:41:53 >11:41:53 Assert Failed: Page Check Error in bttreemrg Page Check error in btree merge?. I would run the above oncheck's. >11:41:53 Who: Session(475, prodadmn@twoface, 24920, 15) > Thread(3414, oncheckm, c5da9578, 1) >11:41:53 Results: Possible inconsistencies in an index of 'rims:"prodadmn".lab >_rslt_a' >11:41:53 Action: Run 'oncheck -cI 5242914' >11:41:53 See Also: /tmp/af.a993b285 >--------------------------------------------------------> >The oncheck for that index produced this error... > >-------------------------------------------------------- > Index lab_rslta_idx02 >Could not check rowids and perform data<->index check >ISAM error: Unexpected internal error Again either a bug or corrupt reserved/extent info. > >Index lab_rslta_idx02 is bad. OK to repair it? >Thu Jan 30 11:41:53 MST 1997 >-------------------------------------------------------- > >Hope to find some solution to this... > >Thanks. > >-- >Yagnesh M Mehta | ymehta@rencorp.com >Gambro Healthcare | (615) 353-4145 > -- David Williams