Secondary adapter instance does not resume publication on check point restart when the Primary instance goes down while running TIBCO ActiveMatrix Adapter for Files (Unix/Win) in FT mode.

Secondary adapter instance does not resume publication on check point restart when the Primary instance goes down while running TIBCO ActiveMatrix Adapter for Files (Unix/Win) in FT mode.

book

Article ID: KB0092926

calendar_today

Updated On:

Products Versions
TIBCO ActiveMatrix BusinessWorks Plug-in for Files for Unix and Windows -
Not Applicable -

Description

Description:

When the TIBCO ActiveMatrix Adapter for Files (Unix/Win) (FA) is configured to run in FT mode, and if primary FA instance stops while it is in the middle of processing a file, then the secondary FA instance becomes active/running from standby but it starts scanning the input directory rather then resuming publication upon check point restart. Ideally, a secondary instance should resume processing the file from the point where the primary instance failed so there will be a complete FT. 


Symptoms:
If the primary file adapter instance stopped during processing a file, then the input file is seen in the working directory. Also, the secondary file adapter will not process this input file and scan the input directory. A configured progress directory of TIBCO ActiveMatrix Adapter for Files (Unix/Win) (FA) instance will have multiple progress files.

Cause:
As the Progress file is being created with the name _[ApplicationName]_[FAServiceName].prg, it is unique for FA instances running in FT which causes this issue.

Issue/Introduction

Secondary adapter instance does not resume publication on check point restart when the Primary instance goes down while running TIBCO ActiveMatrix Adapter for Files (Unix/Win) in FT mode.

Environment

Product: TIBCO ActiveMatrix® Adapter for Files for Unix/Win Version: 7.0.x OS: All Supported Operating Systems --------------------

Resolution

Make sure that the TIBCO ActiveMatrix Adapter for Files (Unix/Win) (FA) instances, which are running in FT, share the same progress file. While deploying the FA application in FT, Admin will change the instance name to force it to be unique. Check the instance name in the deployed tra file and change it to have the same value for all instances running FT for FA, i.e., keep the same instance name in the deployed .tra file of the primary instance and secondary instance.

Manually setting tibco.instanceid with a common value in FA instances running in FT helps to resolve the issue. Modify the "tibco.instanceid" property in the deployed application.tra in such a way that the values are the same for FA instances running in FT.

Additional Information