BusinessConnect Poller Management
Article ID: KB0093184
Updated On:
| Products | Versions |
|---|---|
| TIBCO BusinessConnect | - |
| Not Applicable | - |
Description
Starting with BusinessConnect 6.1.X, the BusinessConnect engine implements a set of microagents to monitor and manage the various polling operations that happen inside the BC engine.
Here are the microagent names and functions:
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=RESENDPOLLER: Poller to look for transactions that need to be resent.
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=QUEUEACTIONPOLLER: Poller to look for messages to send off the message queue (primarily batch transactions)
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=MDNPOLLER: Poller to look for overdue MDNs with AS2
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=InteriorController: The main BC engine
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=HIBERPOLLER: Poller to look up transactions that have hibernated too long
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=FTPPOLLER: Poller that triggers inbound FTP inbound sessions on a periodic basis. This poller is vulnerable to hanging, so it would be good to monitor it.
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=CREDPOLLER: Expired credential poller. Checks certificate store for expired certicates on a regular basis.
BusinessConnect-FL-TIBCODEV-VM-Interior-Server:name=CANCELPOLLER: Poller that looks at audit log for transaction that have been marked for cancellation. Primarily for RosettaNet and ebXML.
For each poller, there are these methods:
| lastCompleteDate | Interior-Server:name=[XXXXPOLLER]:_getlastCompleteDate() |
| isPotentiallyHung | Interior-Server:name=[XXXXPOLLER]:_getisPotentiallyHung() |
| isHung | Interior-Server:name=[XXXXPOLLER]:_getisHung() |
| destroyTheHangingThread | Interior-Server:name=[XXXXPOLLER]:destroyTheHangingThread() |
Refer to "Appendix D Application Monitoring and Management by Processing Rulebases" for information on how to utilize these Microagents.
The most common reason for monitoring BC is for the FTP pollers. In an environment where there may not be reliable data connections to trading partners, there is a tendency for this poller to "hang" because of a bad connection to a SFTP trading partner. Using microagents, you can eliminate this issue with a Hawk rule and rulebase.
Issue/Introduction
Additional Information
Feedback
