turbo log error

This is a segment of my turbo.log. Running Turbo 1.10.03M on Unisys 5000. Users slowly began to come to halt. Turbo Off and then re-starting fixed the problem. tbstat at the time showed daemon no longer running. Since no-one kills them, why do they die? Does the daemon dying cause these error messages or does whatever caused the messages then kill the daemon also? Is there anything in 4gl programs that would cause this to happen? 11:04:22 Checkpoint Completed 11:17:36 Process Aborted Abnormally: pid = 3885, user = 421, flags = a201 11:17:36 ERR: logundo: type = 11, iserrno = 0 (more of these...) Mon Feb 8 11:38:01 1993 11:38:01 Turbo Initialized -- Shared Memory Initialized 11:38:02 Physical Recovery Started 11:38:06 Physical Recovery Complete: 32 Pages Restored 11:38:07 Logical Recovery Complete: 11 Committed, 0 Rolled Back 11:38:11 Checkpoint Completed 11:38:11 Quiescent Mode 11:38:14 On-Line Mode 11:43:12 Checkpoint Completed ...4GL error logs showed only some of these.... Date: 02/08/93 Time: 11:16:00 Program error at c1invlde.4gl line 1011 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Date: 02/08/93 Time: 11:16:46 Program error at c5dpsuiq.4gl line 55 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Date: 02/08/93 Time: 11:19:40 Program error at c8rlkddb.4gl line 42 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Later in the same day the users again began to slowly die. Daemon was still running and no such messages as these appeared. But again we had to turn turbo off and re-start to get users back in. tbstats at that time showed everyone waiting for a latch (lock)? S-B.... beside their name in tbstat -u. No error messages from the 2nd occurrence, but there is a core dump in the directory with the file where the programs write their errors. strings on core shows yacc stack overflow. Is this significant? (I know nothing of core files or how to interpret them!) Turbo.log.... 13:29:55 Checkpoint Completed 13:30:56 Process Aborted Abnormally: pid = 18321, user = 103, flags = a201Mon Feb 8 14:20:41 1993 14:20:41 Turbo Initialized -- Shared Memory Initialized 14:20:41 Physical Recovery Started 14:20:43 Physical Recovery Complete: 64 Pages Restored 14:20:45 Logical Recovery Complete: 15 Committed, 0 Rolled Back 14:20:46 Checkpoint Completed 14:20:46 Quiescent Mode 14:21:11 On-Line Mode 14:36:35 Checkpoint Completed 14:41:54 Checkpoint Completed 4GL messages....(not in chornological order, because of rollbacks??) Date: 02/08/93 Time: 13:17:23 Program error at t2empdat.4gl line 315 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Date: 02/08/93 Time: 13:28:32 Program error at t2settrt.4gl line 687 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Date: 02/08/93 Time: 13:18:43 Program error at c1copo01.4gl line 55 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Date: 02/08/93 Time: 13:58:46 Program error at t2wkldwk.4gl line 226 SQL statement error number -408. ISAM error number 0.Invalid message type received from the sqlexec process. Anything I should fix/tune? We have been using the system for many years, and 2 problems like this in 1 day is most unusual. Could be hardware beginning to go is my first thought. Any ideas? Thanks. BTW, I have found everyone input from this list, flames excepted, to be very helpful. I am learning about problems BEFORE they happen..really useful! Ann Barnes. If email reply is preferred, please write to anbarnes@letterkenn-emh1.army.mil