Posted: 21 Jan 2016 10:02 EST Last activity: 9 Jun 2020 20:02 EDT
JDBC Isolation Level
Does the Transaction Commit Level include a switch for Read Uncommitted transaction? If not, we would like to understand in further detail how to introduce such a change into our implementation of the system
Thank you Vipin. I did that. However, now I don't see any explicit calls to setTransactionIsolation method and now I have to believe based on that.... Pega is using the Default Isolation Level. (read committed) Given that information is now confirmed. How do I go about submitting an enhancement request to Pega Engineering so that I can have some sort of configuration based option in the production. Since it is something which must invoke SetTransactionIsolation in the proprietary Java Code maintained by Pega
Typically users don't use the "Run" button on the Rule-Admin-Extract ruleform to run extracts. This button underneath uses the executeExtract API. They use the command line version of BIX. You can refer to this in the BIX User Guide - https://pdn.pega.com/documents/bix-71-user-guide (This requires the setup of its own data source in prconfig.xml file)
That said, we do not change the isolation level that we get from the connection pool for this API.
Hi Rajiv, that was my point. I need to be able to change the connection setting but I cannot because it is inside a Private API that only Pega internal code has access to and is not configurable anywhere else, like in an App Container such as Websphere. I am implementing Pega BIX as was prescribed through direct invocation from command prompt / shell script.
I can actually take care of the enhancement request here Kevin Zheng, no need to open an SR. I'll update the thread when I have the enhancement request number. Robert Seplin I may reach out to you via private message for additional information if necessary.
So for the command line BIX, would it be possible for you to set up your own connection pool with the required isolation level rather than using the executeExtract API? As I mentioned, executeExtract is not the preferred way as extracts can be resource intensive depending on the size and format of the extract and therefore it is best done via the command line option provided.