[Logo] Terracotta Discussion Forums (LEGACY READ-ONLY ARCHIVE)
  [Search] Search   [Recent Topics] Recent Topics   [Members]  Member Listing   [Groups] Back to home page 
[Register] Register / 
[Login] Login 
[Expert]
Messages posted by: orion  XML
Profile for orion -> Messages posted by orion [96] Go to Page: 1, 2, 3, 4, 5, 6, 7 Next 
Author Message
It works for me and a few other people I talked to today. The kit seems fine and the md5 matches the kit we tested.
That page doesn't exist. May I ask where you found the link to that page?
I've just tested all of the links and I occasionally get an error from AmazonAWS. I'm going to see about putting up a secondary download link in case S3 remains flaky.
It's not possible to make changes to clustered objects while disconnected from the Terracotta server.
We have to update the online Examinator docs to reflect the usage of this. But, you can see it work in the actual code.
We have recently added new documentation that will help you understand how locking in Java and Terracotta works:

http://terracotta.org/web/display/docs/Concept+and+Architecture+Guide#ConceptandArchitectureGuide-Locks
One thing you can do is try to minimize the number of classes that you are instrumenting down to just exactly what you need to be instrumented.

If you've already done that and the delay is still unacceptable, you could have an initial program (if in Java, then it would be in a JVM that isn't clustered) that would show the user a splash screen while it launches the clustered JVM that runs your actual application.

There might be a slicker way to do it, but I don't know what...
Are you trying to cluster http sessions, or your own custom objects?
The number of threads that can run (actually) concurrently is determined by the number of processors/cores your hardware has.
We've recently come down even stronger against using concurrent locks. They really aren't safe to use unless you know exactly what you're doing.

In the case described in this discussion where it is unlikely that different threads will be contending for the lock (except in cases of failover), Kunal is right: write locks will be quite efficient, since, after the first lock award, they will all be awarded locally and, hence, at memory speed.
Which is now resolved.
I think you may be running into this bug:

http://jira.terracotta.org/jira/browse/CDV-792
I think this is the getting started guide Taylor was talking about: http://www.terracotta.org/confluence/display/howto/Getting+Started+with+JBoss+Application+Server
Just so you know, Terracotta does offer a production support package that you might want to consider.

Our subscription service covers:
  • Guaranteed support response times
  • Commercial License with Warranty and Indemnification
  • Bug Fixes in your Production Branch

    and much more.

    Full details of what is contained in the subscription can be found here:
    http://www.terracottatech.com/confluence/display/comsite/Subscription

    Of course, we're happy to help you here on the forums, but the forums support is "best effort" (which basically means we try to be as helpful as we can, but there are no guarantees that your issue will be addressed in a timely manner). If you want a guaranteed response for your production issues, the support subscription is the way to go.

    Cheers,
    Orion
  • unjar it in the eclipse plugins directory. For example, on my mac, it's like this:

    %> cd /Applications/eclipse/plugins
    %> tar xfv <path/to/jar/file>
     
    Profile for orion -> Messages posted by orion [96] Go to Page: 1, 2, 3, 4, 5, 6, 7 Next 
    Go to:   
    Powered by JForum 2.1.7 © JForum Team