Permalink
Commits on Sep 23, 2009
  1. Updated site task.

    dustin committed Sep 23, 2009
  2. Use emma as the coverage tool.

    dustin committed Sep 23, 2009
  3. ConnectionFactoryBuilder

    dustin committed Mar 2, 2009
Commits on Sep 20, 2009
  1. Optimize binary protocol sets.

    dustin committed Sep 20, 2009
    This optimization works in a similar fashion to the GET optimization,
    but is binary protocol specific.  Effectively, all mutation operations
    are folded into quiet ops and the responses are dispatched as implied
    positives when the operations are known to have finished.
Commits on Sep 19, 2009
  1. Added a bulk loader API.

    dustin committed Sep 19, 2009
  2. Refactor optimization control.

    dustin committed Sep 19, 2009
    No longer a dangerous mutable field in the connection.  The
    ConnectionFactory defines it and it's used throughout the life of the
    connection.
Commits on Sep 18, 2009
  1. Provide some breadcrumbs about which nodes were in contact during tim…

    dustin committed Sep 18, 2009
    …eouts.
    
    The root cause of an OperationTimeoutException should have something
    that looks like this:
    
    Caused by: net.spy.memcached.internal.CheckedOperationTimeoutException:
           Timed out waiting for operation - failing node: /127.0.0.1:11213
    
    The actual failing operation is available from the
    CheckedOperationTimeoutException in case someone wants to do something
    particularly special with it.
Commits on Sep 6, 2009
  1. Test for TranscodeService.

    dustin committed Sep 6, 2009
  2. Asynchronous transcoder service.

    Joseph Hart authored and dustin committed Sep 6, 2009
  3. Conditionally allow enabling/disabling the Nagle algorithm.

    pi3orama authored and dustin committed Sep 6, 2009
    Off by default.
Commits on Sep 3, 2009
  1. buildr cobertura update

    dustin committed Sep 3, 2009
Commits on Aug 13, 2009
  1. Handle failures on initial connection.

    dustin committed Aug 13, 2009
    If a ConnectException is thrown while performing the initial
    connection, just queue it and let it reconnect naturally.
    
    Note that initially tried running *all* connections through the queue,
    but it made startup considerably slower
  2. Testing a really long CAS identifier.

    dustin committed Aug 13, 2009
    (issue 80)
Commits on Jun 21, 2009
  1. Allow CAS Mutator to work in a mutation-only mode.

    dustin committed Jun 21, 2009
    When initial is null, do not initialize the value.
Commits on Apr 14, 2009
  1. Fix bug when a reconnect is scheduled simultaneously for two connecti…

    dustin committed Apr 14, 2009
    …ons.
    
    This is a time-sorted map, so adding one could kick another one out.
    As it's rare, detecting it and fudging the reconnection time is the
    least intrusive and safest way to prevent things from getting dropped.
Commits on Apr 2, 2009
  1. Allow transcoders to say how big objects they cache are.

    dustin committed Apr 2, 2009
    This allows someone to, for example, have a transcoder for small
    objects only, or allow access to a server that supports arbitrarily
    large objects.
Commits on Mar 31, 2009