Performance with 4.6C and ECC6

[Mike Gambier]

Rick,
 
Similar picture here as it happens. No Java stack and no intention of using ABAP Classes or XML containers yet. But obviously the Workflow engine has been re-written with these things in mind so yes it's proving hard to comapre in terms of runtime like-for-like.
One thing is absolutely clear though: ECC 6 is a total memory hog by comparison. SAP have advised us to add on 50% more workspace needed to keep the same traffic flowing just to be sure and they don't seem to be far off the mark.
 
We're running full scale volume tests with a Pre-Production environment starting this weekend (rehersal for the real upgrade) so I'll have more data to hand next week. I'm keeping my eye on some of our heavy workflow steps to see how long they take in the new environment. The only drawback is that we won't have exactly the same hardware to play with (4 app servers instead of 16) but at least I should be able to compare the same Workflow instances and spot any odd timing differences between 4.6c and ECC 6.
 
I have taken the precaution to change all of our Workflow definitions to continue to use the 'old' BOR Object-based container (Structure Persistence) for now. You may want to do the same as the default setting (Compatibility) will seek to switch to the new XML container when the release version of the Workflow Header is later than 4.6c. But that's up to you of course :)
 
I've not heard the '30% slower' figure by the way. My gut feel is that things are slightly slower all round, particularly deadlines, but I've not noticed that much of a delay, yet.
 
Mike GT



Subject: RE: Performance with 4.6C and ECC6Date: Fri, 12 Sep 2008 08:54:47 -0500From: Rick.Sample at graybar.comTo: sap-wug at mit.edu



Mike,
 

I am sure Management / BASIS will take the standard recommendations for hardware. So, backend stuff is not concerned for this request. 
 
So, more on the App process. 
Wf starts. Takes X time to process task XYZ, ...
 
We won't have the Jave stack installed from my understanding, so that is out of the picture. 
The only thing that comes to mind on performance is the new XML changes. But all our apps will continue to use the tables vs. XML. 
 
I hear that ECC6 is aprox 30% slower that 4.6C. Don't know how that number keeps coming up, but like comparing apples to oranges me thinks. 
 
Thanks,
Rick
 


From: sap-wug-bounces at mit.edu [mailto:sap-wug-bounces at mit.edu] On Behalf Of Mike GambierSent: Thursday, September 11, 2008 3:34 PMTo: SAP Workflow Users' GroupSubject: RE: Performance with 4.6C and ECC6
Rick, We're doing the same at the moment. Are you focusing on the Basis side of things (e.g. like 'buffer gets' on the db, database locks) or the Application/Business Process area (e.g. event delivery, deadlines etc) or both? Mike GT

Subject: Performance with 4.6C and ECC6Date: Thu, 11 Sep 2008 14:36:21 -0500From: Rick.Sample at graybar.comTo: sap-wug at mit.eduCC: rick.sample at graybar.com
We are preparing to upgrade 4.6c to ECC6. I have been tasked to do a "performance compare" between the two systems for WF relegated items.
 
Anyone have any tools, methods, etc. to do this type of performance checks? I am not sure as to what to compare or how.
 
 
 Rick Sample | SAP Business Workflow DeveloperOffice (314) 573-5822 | Mobile (314) 952-2273 | rick.sample at graybar.com www.graybar.com - Graybar Works to Your Advantage 
 

Get Hotmail on your mobile from Vodafone Try it Now! 
_________________________________________________________________
Get all your favourite content with the slick new MSN Toolbar - FREE
http://clk.atdmt.com/UKM/go/111354027/direct/01/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/sap-wug/attachments/20080912/9da0e3fa/attachment.htm