Skip to content
Commits on Apr 14, 2011
  1. @jashkenas

    CloudCrowd 0.6.2

    jashkenas committed
  2. @jashkenas
Commits on Mar 12, 2011
  1. @jashkenas

    CloudCrowd 0.6.1

    jashkenas committed
  2. @jashkenas
Commits on Mar 10, 2011
  1. @jashkenas
Commits on Mar 7, 2011
  1. @jashkenas

    comment the horrible hack.

    jashkenas committed
  2. @jashkenas
  3. @jashkenas

    fixing SQL syntax.

    jashkenas committed
  4. @jashkenas
Commits on Mar 4, 2011
  1. @jashkenas
  2. @jashkenas
Commits on Feb 17, 2011
  1. @jashkenas
Commits on Feb 16, 2011
  1. @jashkenas
  2. @jashkenas

    mo comments

    jashkenas committed
Commits on Feb 14, 2011
  1. @jashkenas

    switching to remote_storage_prefix and local_storage_prefix to allow …

    jashkenas committed
    …tmpdirs to be fully cleaned up on worker machines.
  2. @jashkenas
Commits on Sep 28, 2010
  1. @jashkenas

    Updating changelog

    jashkenas committed
  2. @jashkenas
Commits on Sep 14, 2010
  1. @felipecvo

    versão 0.5.2

    felipecvo committed
Commits on Sep 10, 2010
  1. @jashkenas
Commits on Aug 5, 2010
  1. @felipecvo

    adding timeout configuration for requests between 'node to server' an…

    felipecvo committed
    …d 'server to node '
    
    Server to node communication is supposed to be fast, always. But node/worker to server can be slow, very slow. So, if you have this issue with timeouts, you can configure how you want this.
Commits on Aug 3, 2010
  1. @felipecvo
Commits on Aug 2, 2010
  1. @felipecvo
Commits on Jul 30, 2010
  1. @felipecvo
  2. @felipecvo
  3. @felipecvo

    If has only one node available and it can't handle work unit action, …

    felipecvo committed
    …break loop and add filter conditions.
  4. @felipecvo

    If node can't handle this action, put it back to available node list,…

    felipecvo committed
    … so it can handle another action.
  5. @felipecvo
  6. @felipecvo
  7. @felipecvo

    After created job, try to distribute work units asynchronously, and r…

    felipecvo committed
    …eturn as soon as possible to client. By this way, when distribution fail or take too long, client request won't get affected.
  8. @felipecvo
Commits on Jun 22, 2010
  1. @jashkenas
  2. @jashkenas

    switching from a curl-shell-out to a streaming net-http for Action#do…

    jashkenas committed
    …wnload. Slower, but more portable.
  3. @jashkenas
  4. @jashkenas
Something went wrong with that request. Please try again.