we include a shell script to run bix extract from outside of application server. Like our other prpcutils script features, we use xml files to configure database connectivity and specify options. The shell fires up a stand-alone jvm that instantiates engine and runs the exportImpl BIX code.
with proper profile and setup, no reason that the shell script can not run under any standard task scheduler, including crontab or CA7
BIX, till 7.1.8, was not available with all the options to be run via a Pega platform agent. If things needed to be scheduled to run at regular intervals, command line (batch script or shell script) was the only option.
Note that when run via an agent, it will run in a background thread in the same JVM as the node where it has been scheduled to run. If run via a shell script, it spawns a new JVM (process) to extract the data. So the call is whether to load the current JVM with extra processing or to spawn a new one.
And to Add to the above conversation, to run BIX from unix script, Pega (application) need not be up and running, BIX will interact with Pega database and perform the extract. No Additional load on Pega application. Please correct my understanding if it is wrong.
Yes you are correct. As Andy stated earlier when running BIX from the command line it fires up a stand-alone jvm that instantiates engine and runs the exportImpl BIX code. There is no interaction with the app server or Pega application. It is strictly dealing with the database. Of course it is the same database that the Pega application is interacting with if it is up and running so you still do not want to be running BIX extractions during peak hours.