using IBM zEnterprise Data Compression on PEGA apps running under java-7
We are trying to use zEDC tool for data compression so we are looking for some inputs on where compression takes place within the PEGA framework… some possibilities were thought to be during rules refresh, or after deleting the pega_rules marker? It seems from the logs that inflate is used a lot, which also supported in the zEDC compression accelerator (not just deflate).
I am adding this to clarify the question as I am interested in the potential CPU offload, and performance boost this may provide.
Section 8.1 / pg 143 in this redbook: http://www.redbooks.ibm.com/redbooks/pdfs/sg248259.pdf explains the Java specific methods eligible for zEDC hardware assistance. Specific code examples are provided. Essentially the question of where does Pega perform deflate/inflate (stowing LOBs?, passivation in filesystem?) and if so, are the default buffers sizes large enough (4k min) to be eligible for hardware offload. If Pega defaults do not meet min buffer size requirements, can they be set via a property?