• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: load testing
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: load testing


  • Subject: Re: load testing
  • From: Chuck Hill <email@hidden>
  • Date: Tue, 26 Jul 2005 10:02:01 -0700

A week later... :-)

On Jul 21, 2005, at 3:34 PM, Mike Schrag wrote:

So I've been load testing my app w/ Apache JMeter while profiling it with JProfiler, both really cool apps that are working really nicely together -- just a public service announcement to start things off.

So single "user" things were great ... Then I tossed 50 concurrent "users" at it and things didn't go so well. CPU usage was relatively low, memory was no problem at all, so obviously I'm thinking lock contention. So I attach JProfiler and sure enough my app is spending a HUGE amount of time (like >30%) in sharedEditingContext.unlock(). Odd because I don't actually use shared editing context. So for grins I added EOSharedEditingContext.setDefaultSharedEditingContext(null) to the top of my app and performance was around 10x faster.

So my three part question:
1) does the original behavior sound odd? it sure SEEMS odd to me -- particularly since nobody else has mentioned this ... which leads me to believe it's my fault ... which is historically a safe bet to make when it comes to Me vs WO


Yes, this seems odd to me too. Was your load testing resulting in more concurrent saveChanges() calls than the app would really encounter? I've never timed this with and without the shared EC. It could be that nobody else has either. Perhaps I can get a huge performance boost!


2) what horrors will i bring upon my family by completely removing the shared editing context ... my tests have not revealed a problem yet, but i don't want to set myself up for trouble

If you are not using it then there is no need for it.


3) is there a better fix for whatever it is I am seeing with this lock contention, or if nobody is seeing this issue, I'll go back to the profiler drawing board and see if i can track down more specifically where the contention is coming from

If you have not found out anything else you should file this on http://bugreport.apple.com

Chuck


--
Practical WebObjects - a book for intermediate WebObjects developers who want to increase their overall knowledge of WebObjects, or those who are trying to solve specific application development problems.
http://www.global-village.net/products/practical_webobjects





_______________________________________________ Do not post admin requests to the list. They will be ignored. Webobjects-dev mailing list (email@hidden) Help/Unsubscribe/Update your Subscription: This email sent to email@hidden
  • Follow-Ups:
    • Re: load testing
      • From: Mike Schrag <email@hidden>
References: 
 >load testing (From: Mike Schrag <email@hidden>)

  • Prev by Date: Re: EditingContext problem, ghost entry in datastore [ was Debuging WOManyToManyRelationship]
  • Next by Date: Re: SSDD vs WAR
  • Previous by thread: load testing
  • Next by thread: Re: load testing
  • Index(es):
    • Date
    • Thread