Processing Rule mismatch in Clustered Environment

We are running webMethods 6.5 in a clustered environment. We have 2 IS, 2 TN and the TNs are sharing the same database. So from I understand, the Partner Profile, Document Type, and Processing Rule on the 2 TNs should be identical and if you make changes on one, it should reflect the changes on the other.

We ran into problems when we export the Partner Profile from DEV and import them into the clustered QA. For some reason, things look fine on the first server but on the second server the Sender or Receiver on the processing rule would look blank even though we chose the right partner profile. See attachment.

I’ve read through a couple threads in the forum and found out about the internal ID between the Partner Profile, Processing Rule, etc etc. But even after creating brand new Partner Profile and Processing Rule, it is still giving that inconsistent behaviour between to two TN.

Does anyone have any idea? Or worked with Clustered IS & TN before?

Any help is appreicated!

The IS cluster doesn’t have an impact on this per se.

For the two TNs to coordinate their sharing of the DB, they notify each other when one of them changes the DB for profiles, rules, etc., as you point out. It will only do this, however, if the TN config says to do so. The applicable properties in properties.cnf are:


These are described in the TN User’s Guide and in the TN online help. You may want to verify that these are configured.

That said, I’m not sure if an import action causes notification to happen. You may need to force a cache reload manually. I can’t remember the service off-hand

We also faced a similar situation in 6.1. But the solution for it what raemon has described above.
Once you have set the properties mentioned above the cache is automatically refreshed on the second if you do any changes on first server

Importing the partner does cause a notiifcation and cache is refreshed on both servers. We use the same setup in 6.1 and it works fine.