Jonathan Leffler/Informix HELP!

I have a problem with database corruption on a Sun Ultrasparc, the client is planning to move their "live" database to this machine in around 2 weeks. Everything was OK until the client loaded some unrelated applications onto 5 pre-existing filesystems. Now it appears we have database corruption with NO overlapping UNIX partitions and NO overlapping Informix chunks:- $ tbstat -VRSAM Version 5.07.UC1 $ uname -aSunOS xxx 5.5 Generic sun4u sparc SUNW,Ultra-1 isql yyy 311: Cannot open system catalog (systables). 135: ISAM error: tblspace does not exist $ tbcheck -cc Validating database yyy ISAM error: tblspace does not exist tbcheck -cr find no errors. If I then run:- $ tbcheck -ce Validating extents... Segmentation Fault - core dumped $ adb $INFORMIXDIR/bin/tbcheckcore file = core -- program ``tbcheck'' on platform SUNW,Ultra-1 SIGSEGV: Segmentation Fault $c wrtppextents(0xa6130,0x0,0x1002d9ac,0x2,0xeffff7d4,0x201f48) + 10 doroot(0xa6130,0x9e6d4,0x1534,0xeffff840,0x0,0xa6130) + 198 chk_extents(0x9b000,0x4,0x0,0xae795,0x0,0xad5e8) + e4 main(0x2,0xeffffb64,0xeffffb70,0xa5800,0x1,0x0) + 304 15:45:52 Process Aborted Abnormally: pid=3909 user=0 us=10001da4 flags=81 15:45:52 Process Aborted Abnormally: pid=3910 user=0 us=10001e38 flags=81 ------------------------------------------------------------------------ --- 16:57:39 Shutdown Mode 16:57:39 Quiescent Mode 16:57:41 INFORMIX-OnLine Stopped Tue Aug 20 16:57:46 1996 16:57:46 bfcheck: bad page: pg_addr 0 != bp->bf_pagenum 300004, userp =10001be 8, pid = 4200, uid = 503 16:57:46 buffer header: 16:57:46 100b5e90: 00000000 00000000 00000000 100b5e10 ........ ......^. 16:57:46 100b5ea0: 100b47e0 10000470 100b5ed0 00020000 ..G....p ..^..... 16:57:46 100b5eb0: 010c0000 00300004 100bf800 00000000 .....0.. ........ 16:57:46 100b5ec0: 00000000 00010000 100b8df0 01000000 ........ ........ 16:57:46 page header: 16:57:46 100bf800: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 100bf810: 00000000 00000000 ........ 16:57:46 slot table and stamp: 16:57:46 100bfffc: 019dec41 ...A 16:57:46 tblspace header error: 16:57:46 1002da10: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da20: 00010000 03000001 00300004 00000000 ........ .0...... 16:57:46 1002da30: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da40: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da50: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da60: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da70: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da80: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002da90: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002daa0: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002dab0: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002dac0: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002dad0: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002dae0: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002daf0: 00000000 00000000 00000000 00000000 ........ ........ 16:57:46 1002db00: 00000000 100ac758 01a00781 00000000 .......X ........ 16:57:46 1002db10: 100b05d8 .... 16:57:46 -- Fail Consistency Check -- pthdrpage:ptalloc:bad partn page -- pid=4200 user=503 us=10001be8 16:57:46 Cannot Open Dbspace 3 16:57:46 INFORMIX-OnLine Initialized -- Shared Memory Initialized 16:57:46 Physical Recovery Started 16:57:46 Checkpoint Completed 16:57:46 Physical Recovery Complete: 0 Pages Restored 16:57:46 Logical Recovery Complete 0 Committed, 0 Rolled Back, 0 Open, 0 Bad Locks 16:57:46 On-Line Mode 17:01:36 Process Aborted Abnormally: pid=4228 user=503 us=10001da4 flags=81 17:19:12 Process Aborted Abnormally: pid=4316 user=503 us=10001da4 flags=81 Informix Tech. Support just said restore from backup and move the root dbspace somewhere else - onto a cooked filesystem. If I restore then I risk corrupting other applications which are not my companies. Surely there must be a way to find out what is overlapping? The engineer was not surprised that tbcheck seg faulted! Surely this is a bug in tbcheck??? All chunks are flags as PO i.e. online surely this is wrong if dbspace 3 cannot be opened?? Jonathan Leffler (or someone else from Informix) HELP! I'm stumped. -- David Williams