5 replies to this topic

[ashumway]

We've got a batch job that runs every minute or so, to transfer data from a non-Pro IV Oracle DB to Glovia. This batch job runs some Pro IV code that checks some Pro-ISAM files and then does a SQL call to use a stored procedure to move the data from one DB to the other. About once per day, we get a Dr.Watson error (access violation). I checked with Microsoft and they say it's caused by Pro IV.
I've tried using exception handlers in the stored procedure, but haven't gotten any result there, so it doesn't appear to be caused by the SQL code. Does anyone have a ideas about how I could track this down in Pro IV? The error doesn't happen every day and I can't force it to happen.
We're running Pro IV version 4.6 on Windows 2000.

[Phil]

It could be elementary Watson.

[ashumway]

God-
Thanks for the suggestion. I guess I don't understand, because the function does not print anything.

[Kevin Bruss]

Can you show us the actual error message from Watson?

Could it could be that perhaps you are running cleanup tasks that are causing record locks, or you have handlers that disable access temporarily while the hourly task is
trying to get in?

If you're using SQL Server (you mentioned Microsoft) perhaps you could run a trace for
a suspected time frame.

I beleive you can also do a trace in oracle as well.

In addition, your task in pro-iv could write to a flat file (just use an 'echo' statement in a system call.. I know there is a discussion on this usage, but.. its only temporary).
Write a 'begin' and 'end' trail to the log with date/time stamps and see if you can catch
the error in between.. perhaps you will only have one record at one point that could illustrate the first notion.

[ashumway]

Kevin,

As far as the traces are concerned, I'm using an ONSERVERERROR trigger to try to trap any Oracle error that might be causing it, but so far I've had no success. In fact, the trigger fires, but doesn't write any data to the table that I'm using because the database session dies first.
Yes, I could post the message, but it's quite lengthy. Here is some of it - I left out the stack dumps:

Application exception occurred:
App: (pid=1524)
When: 3/1/2004 @ 12:09:24.907
Exception number: c0000005 (access violation)


*----> Task List <----*
0 Idle.exe
8 System.exe
140 smss.exe
164 csrss.exe
184 WINLOGON.exe
212 services.exe
224 lsass.exe
404 svchost.exe
432 SPOOLSV.exe
488 svchost.exe
528 FrameworkServic.exe
584 NMSSvc.exe
624 regsvc.exe
652 rshsvc.exe
668 r_server.exe
700 mstask.exe
776 naPrdMgr.exe
796 winmgmt.exe
836 svchost.exe
848 PSXRUN.exe
864 PSXSS.exe
872 mapsvc.exe
1124 zzInterix.exe
1164 init.exe
1100 inetd.exe
1104 cron.exe
576 explorer.exe
320 sistray.exe
1080 KHOOKER.exe
548 PROMon.exe
252 UpdaterUI.exe
1412 svchost.exe
1436 POSIX.exe
1504 ksh.exe
1216 ksh.exe
1524 proiv.exe
1196 drwtsn32.exe
0 _Total.exe

function: Ordinal5085
5f402d91 33c0 xor eax,eax
5f402d93 85c9 test ecx,ecx
5f402d95 c745d804c4495f ss:00ba9c26=????????
mov dword ptr [ebp+0xd8],0x5f49c404
5f402d9c 8975ec mov [ebp+0xec],esi ss:00ba9c26=????????
5f402d9f 894df8 mov [ebp+0xf8],ecx ss:00ba9c26=????????
5f402da2 8945e0 mov [ebp+0xe0],eax ss:00ba9c26=????????
5f402da5 7638 jbe Ordinal5198+0x71 (5f40b8df)
5f402da7 8b4e60 mov ecx,[esi+0x60] ds:019d2dae=????????
5f402daa 8d0480 lea eax,[eax+eax*4] ds:0000000a=????????
5f402dad 6a00 push 0x0
FAULT ->5f402daf 8b0481 mov eax,[ecx+eax*4] ds:0000000a=????????
5f402db2 8d4dd8 lea ecx,[ebp+0xd8] ss:00ba9c26=????????
5f402db5 51 push ecx
5f402db6 6aff push 0xff
5f402db8 50 push eax
5f402db9 8bce mov ecx,esi
5f402dbb 8945dc mov [ebp+0xdc],eax ss:00ba9c26=????????
5f402dbe e8b1f3ffff call Ordinal4424 (5f402174)
5f402dc3 85c0 test eax,eax
5f402dc5 750c jnz Ordinal5122+0x4f (5f407dd3)
5f402dc7 50 push eax
5f402dc8 8d4dd8 lea ecx,[ebp+0xd8] ss:00ba9c26=????????