btree page check error

IDS 10.00.UC4, RedHat AS 3 Got the following ALARM messages about an index at 07:41: SEVER: 4 CLASS: 2 MSG1: Index failure: 'spirit_beta_db:resvis.book_fares_segm#d2book_fares_segm'. MSG2: Page Check Error in btc_scan_next:bad btree leaf scan_next SEVER: 5 CLASS: 6 MSG1: Internal Subsystem failure: 'MT' MSG2: Page Check Error in btc_scan_next:bad btree leaf scan_next This is from an af.* file dumped at the time: 02:21:18 02:21:18 IBM Informix Dynamic Server Version 10.00.UC4 Software Serial Number AAA#B000000 02:21:18 Assert Failed: Page Check Error in btc_scan_next:bad btree leaf scan_next 02:21:18 Who: Session(57, root@resbe1, 0, 0xc38ccd10) Thread(115, btscanner_0, c38af0c0, 1) File: rsdebug.c Line: 1066 02:21:18 Results: Possible inconsistencies in index 'spirit_beta_db:"resvis".book_fares_segm#d2book_fares_segm' 02:21:18 Action: Run 'oncheck -cI spirit_beta_db:"resvis".book_fares_segm#d2book_fares_segm' 02:21:18 Stack for thread: 115 btscanner_0 base: 0xc4698000 len: 69632 pc: 0x08786a17 tos: 0xc46a8850 state: running vp: 1 0x08786a17 (oninit)afstack (0xc4690438, 0x11af, 0xc46a8960, 0x42a, 0x0, 0x1) 0x087861b9 (oninit)afhandler(0x2, 0x8c44960, 0x8c453c0, 0x8c436a0, 0x401, 0x1) 0x0878565b (oninit)affail_interface(0x8c44960, 0x8c453c0, 0x8c436a0, 0x1, 0x8ae6844, 0x42a) 0x085999fd (oninit)bffail (0x10, 0x8af5588, 0x0, 0xc38af0e0, 0xc38af3e4, 0x0) 0x0864dc04 (oninit)btc_scan_next(0xc40a68e4, 0xc46a8f58, 0xc505d3e0, 0xc40a6900, 0xc40a76ec, 0xc505d3e0) 0x0864d102 (oninit)btc_leaf_scan_index(0x1, 0x0, 0x34, 0xc46a8fa8, 0xc40a7724, 0xc4690438) 0x0864cd26 (oninit)btc_scan_index(0x0, 0x0, 0x0, 0x0, 0xc4690438, 0xa) 0x08649bcb (oninit)btscanner_loop(0xc40a68a8, 0xc4690438, 0x0, 0x0, 0x0, 0x0) 0x08767514 (oninit)startup (0xe, 0xc46abf10, 0xc46b2fd0, 0x2f10, 0x88d61df1, 0x0) 0x00000000 (*nosymtab*)0x0 At the time of the ALARM, this is all that is in the online log: 07:39:31 Maximum server connections 74 07:40:36 Logical Log 303952 Complete, timestamp: 0xab11fb28. 07:40:39 Logical Log 303952 - Backup Started 07:40:54 Logical Log 303952 - Backup Completed 07:41:25 Page Check Error in btc_scan_next:bad btree leaf scan_next 07:41:34 Checkpoint Completed: duration was 2 seconds. 07:41:34 Checkpoint loguniq 303953, logpos 0x321160, timestamp: 0xab129087 I'm confused. A 'page check error' occurs at 02:21 and no ALARM. Then, at 07:41, a one-line error message in online.log and ALARMs go off. Not to mention, I run the 'oncheck -cI' and everything shows ok. Did the engine fix the bad btree? I've heard anecdotal evidence that sometimes a slow I/O system will trigger btree/buffer errors. Has anyone else seen this before? Bob Roussey