Error Request for Progress Review – ipurch DB Abnormal Shutdown with WebSpeed Memory Violation and APW/AIW Lock Stack Errors

[Mike]

The probability that the cause is rcode corruption is, IMHO, very low. what is IMHO very low?Are these rcode developed by applicaiton team or default?

 

[George Potemkin]

Are these rcode developed by applicaiton team or default?

src/web/objects/web-disp.p is a standard program but your application calls it from $DLC/src. It can be a customized version.
tptstart-new.p is a part of application.

 

[jdpjamesp]

You're running Webspeed. It looks like these are shared memory Webspeed clients. The symptoms you are seeing are looking incredibly similar to ones I was experiencing.

The cause of mine was a program that was tidying up handles for persistent procedures. It walked the widget tree and was finding handles for persistent procedures it didn't own and was deleting them. This caused symptoms VERY similar to the ones you are using. The solution is to upgrade to the latest service pack. I'm not sure how many more times we should need to say this. It's essential. It's not just recommended. The other part of this is to make sure that programs that create persistent procedures also delete them. This is a code issue. The FINALLY statement can be used to clear up things in all cases, including error states (which was what caused our issues).

Slight aside: Are you using PDF Include by any chance? We are and that was what caused the errors - the PDF Include process wasn't being cleared up when a program errored.

 

[George Potemkin]

This issue is occurring only with this database. We have around 12 databases running on the same server with the same OpenEdge version, and they are not experiencing any similar issue.

Does each client connect to all these databases?
Is ipurchprod the first database the clients connect to?