When trying to start several bwengines in one shot either with the option automatic restart on boot or manually thru admin, some of the bwengines fail with "NullPointerException ... RepoNodeVFileImpl"
Article ID: KB0087001
Updated On:
| Products | Versions |
|---|---|
| TIBCO Administrator | - |
| Not Applicable | - |
Description
Description:
When trying to start several bwengines in one shot either with the option automatic restart on boot or manually thru admin, some of the bwengines fail with the below errors. The following are the configuration details
1. Domain is configured to use Oracle database
2. Domain database is configured to have Min Connections = 0 and Max Connections=5
3. BWEngines are deployed using RV method of deployment, with "start on boot" option enabled.
4. 11 bwengines are attempted to start in one shot on reboot of a machine.
Startup behavior is inconsistent but randomly some processes initialize correctly on startup; many say they are started, but have not initialized correctly.
1) nulljava.lang.NullPointerException at com.tibco.objectrepo.vfile.tibrepo.RepoNodeVFileImpl.(RepoNodeVFileImpl.java:50) at com.tibco.objectrepo.vfile.tibrepo.RepoNodeVFileImpl.(RepoNodeVFileImpl.java:31) at com.tibco.objectrepo.vfile.tibrepo.RepoObject.(RepoObject.java:49) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.a(RepoVFileFactory.java:1661) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.a(RepoVFileFactory.java:1728) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.if(RepoVFileFactory.java:1673) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.migrateToStream(RepoVFileFactory.java:1461) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.migrateToStream(RepoVFileFactory.java:1388) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.initClient(RepoVFileFactory.java:174) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.(RepoVFileFactory.java:89) at com.tibco.objectrepo.vfile.tibrepo.RepoVFileFactory.(RepoVFileFactory.java:76) at com.tibco.pe.core.RepoLoader.(RepoLoader.java:88) at com.tibco.pe.PEMain.if(PEMain.java:272) at com.tibco.pe.PEMain.a(PEMain.java:157) at com.tibco.pe.PEMain.(PEMain.java:144) at com.tibco.pe.PEMain.main(PEMain.java:82)
2) error creating clientServer not responding com.tibco.infra.repository.OperationFailedException
Symptoms:
If you have larage number of BW application configured to "Start On Boot" we have observed few BW apps will have Startup error.
Cause:
Low Max DB Connection configured , we need large value to support large no of BW application and concurrent users.
Resolution
1. Increase the domain database connection settings to suit to the number of concurrent users. In this case the connections are increased to Min Connections = 15 and Max Connections=50, then the restart of bwengines was very quick and consistently successful.
2. Change the deployment mode to LOCAL. When deployment mode is changed to LOCAL, dat file of the application is stored on the machine where the bwengine is deployed to run. This will eliminate the lot of communication delay that takes place, during startup of the bwengine, due to communication between admin and databse to retrieve the dat file and load it.
2. Change the deployment mode to LOCAL. When deployment mode is changed to LOCAL, dat file of the application is stored on the machine where the bwengine is deployed to run. This will eliminate the lot of communication delay that takes place, during startup of the bwengine, due to communication between admin and databse to retrieve the dat file and load it.
Issue/Introduction
When trying to start several bwengines in one shot either with the option automatic restart on boot or manually thru admin, some of the bwengines fail with "NullPointerException ... RepoNodeVFileImpl"
Feedback
Yes
No
Powered by
