Oengine error (OpenInsight 32-bit Specific) [Revelation On-Line Wiki]

Sign up on the Revelation Software website to have access to the most current content, and to be able to ask questions and get answers from the Revelation community

At 23 JUL 2002 09:00:06AM Michael Norton wrote:

OpenInsight 32-bit Specific

Hi all

I am running a routine that reads data from a dos file extracts a key value and check if record exists on OI file. If it does I null attribute 3 and write it back. Each record is 4 attributes in length.

After processing almost 85% of the dos file I get the following error:

OENGINE caused an invalid page fault in

module LHCLNT.DLL at 0167:21004a85.

Registers:

EAX=00000000 CS=0167 EIP=21004a85 EFLGS=00010246

EBX=015c3c5f SS=016f ESP=0067eb08 EBP=0067f7f4

ECX=0067ecd4 DS=016f ESI=02c4fba8 FS=3d87

EDX=0067eb74 ES=016f EDI=0067ecd4 GS=0000

Bytes at CS:EIP:

c7 00 00 00 00 00 c3 8b 51 04 5e 89 02 8b 51 04

Stack dump:

21004a66 02c4fba8 0067ecd4 0067ecc8 0067eb50 2100d269 02c4fba8 0067ecd4 015c3c69 0067eb50 2100d131 0067eb50 0067ecc8 2100ba5b 0067eb50 00000000

When I manage to restart OI I check the OI file for GFE's and there is an error on frame 0 which is 'free frame list is not sorted'.

Running routine on local copy of routine.

Started from scratch a few times and it falls over in the same place each time.

whats going on!!!!!!!!!!

At 23 JUL 2002 11:07AM Pat McNerthney wrote:

Micheal,

Ouch! Looks like a bug in the depths of Linear Hash.

Is there any chance you could create an RDK with everything required to reproduce this bug? I'd really like to get to the bottom of this one.

Pat

At 26 JUL 2002 09:38AM Michael Norton wrote:

Pat

sorry for delay but been on the road.

I tried running the routine against data from 6 months ago. It is the same file but a copy from 6 months ago and it ran fine.

I then compressed the current file, which had the error, in database manager and ran the routine again it worked file this time.

Does that change your view of what may have caused the error

Cheers

At 26 JUL 2002 12:48PM Pat McNerthney wrote:

Michael,

Apparently there is a bug in LH that will GPF while processing your "corrupted" file, which doesn't surprise me.  If after compressing the file, which essentially rebuilds it, everything is fine, I will stop worrying about it.

Cheers!

Pat

At 26 JUL 2002 12:58PM Oystein Reigem wrote:

Neat. Compress the table and the bug gets squashed. I think I must compress my SYSPROCS right away.

- Oystein -

View this thread on the forum...