Guess that was the issue I was interested in. We're a small company and can't really afford a bunch of n way machines for processing. Some of our customers have these huge servers and expect us to handle thousands of transactions per minute. We've found quite a few tweaks, both in our applications and the DB2 and Oracle back end databases. I'm just looking (OK, thinking with my mouth open) to see if farms are more cost effective than our local hardware. Thanks everyone for your input. Might also be an effective way to do a customer presentation. --- Wayne Johnson, | There are two kinds of people: Those 3943 Penn Ave. N. | who say to God, "Thy will be done," Minneapolis, MN 55412-1908 | and those to whom God says, "All right, (612) 522-7003 | then, have it your way." --C.S. Lewis ________________________________ From: Josh Paetzel <josh at tcbug.org> To: Wayne Johnson <wdtj at yahoo.com> Cc: Elvedin Trnjanin <trnja001 at umn.edu>; steve ulrich <sulrich at botwerks.org>; Mike Miller <mbmiller+l at gmail.com>; TCLUG List <tclug-list at mn-linux.org> Sent: Tuesday, July 7, 2009 1:27:14 PM Subject: Re: [tclug-list] farms for stress testing (was: Re: cheapest "farm"?) On Jul 7, 2009, at 12:57 PM, Wayne Johnson wrote: We have a need to stress test our product. We have a few multi core machines to run as a DB server and App server but they are pretty heavily used and hiding under a desk. > >Now the question. Would it be reasonable to try and run stress testing on EC2 (or other) farms? Since we only will need them occasionally, but beat them to death when we do? Would it be cost effective to run this on a farm? If there is no control over how much memory bandwidth you get, you may not be able to get a consistent load. Is there a similar issue with disk I/O? > > > I guess it all depends on what you are really testing for, and if the differences between testing it on EC2 and it's actual use case really matter or can be factored out. Id be fairly reluctant to think I could get meaningful results for anything approaching actual use cases. Like for instance, what does our application do when 400 people log in might be completely different between EC2 and real deployments....and the kicker is, no matter what it does in testing, you aren't really going to to know what it does in production until you accurately simulate production. There are other sorts of things that you might find with EC2 though, where deploying on real hardware will be overkill. Like for instance how much CPU time does our forking model consume on 8 way hardware with 40 simultanious transactions going on. Thanks, Josh Paetzel -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20090707/dceacda1/attachment.htm