Application Ruleset Stack - Understanding Impact on Application performance
Hi,
In our application we have overall 23 unique rulesets... as part of one major application enhancement ...for some 11 main rulesets corresponding branch rulesets were created (in v5.5) to do project specific coding (similar to branching in v7).... since there were several issues reported in integrated applications... Project team decisded to release this as beta version to validate production usage. So to achive this, project team added those branch rulesets .. on top of existing ruleset in application rule (they were never merged to it's main rulesets).. so now application have 34 unique rulesets.
This is 4 years back story..:) later all development contunue happening in branch ruleset itself.... now this branch rulesets contains major chunk of rules.
We are looking to cleanup this mess by skimming branch ruleset and moveing those skimmed rules back to it's original ruleset, so that ruleset stack can be kept at optimal count.
Now to propose this to client, we have to document it's benifits and a way to measure the impact. Can some one clarify what beinfit we get keeping unique ruleset count to optimal and is there a way to measure pre and post application stats w.rt performance....
***Updated by Moderator: Marissa to update categories***
Hi Asif,
thanks for reply... We are aware of ruleset maintenance wizard details.
My question was if application have 30+ ruleset and if we consolidate them to 10-12 rulesets.... what is the major benefit we get. We know it looks clean and maintenance will be easy., less clutter.. so on..
Apart from this what else is major benefit.. is it improves system performance or educe cache sizes ... and so on?
Hi,
Benefit of skimming:-
- It speed up the PRPC rule resolution process as the old rules aren't considered.
- It de-load the pr4_rules table, thus database interaction gets optimized.
- Quicker deployments due to small packages.
- Easy maintenance.
Burder of skimming:-
- History table gets more entry and overloaded.
- Regression/sanity test of application.
Kindly notify if it helps.
Regards,
Asif
Go to SMA Advanced -> Reports -> Most Recently Used Cache Report or Rule Assemble Cache Management Report. See if any cache is getting pruned or not.To Avoid cache pruning, it is good practice to keep the cache size within the defined limit(Though limits can be defined to higher value than the default value of 20000). And more the rules/Rulesets, bigged is the cache stack. This gives you a reason to limit the Rulesets and skim that.
Hi MGR,
Thanks for posting your query.
Use the RuleSet Maintenance wizard to perform routine RuleSet tasks, such as skimming or compressing RuleSet versions, merging several RuleSets into one, and deleting unneeded RuleSets.
Refer :- https://pdn.pega.com/how-skim-compress-merge-copy-and-delete-rulesets-r…
How to refactor ruleset versions :- https://pdn.pega.com/how-refactor-ruleset-versions
I hope above links helps you. Kindly notify if it does.
Regards,
Asif