Permalink
Commits on Oct 28, 2010
  1. added initial byteBuffer conversions

    zznate committed Oct 28, 2010
Commits on Oct 26, 2010
  1. more intelligent queue polling fixing potential dangled thread issue,…

    zznate committed Oct 26, 2010
    … updated pom rev
Commits on Oct 25, 2010
Commits on Oct 21, 2010
  1. updated snapshot date

    zznate committed Oct 21, 2010
Commits on Oct 19, 2010
Commits on Oct 18, 2010
  1. upped rev number for snaphost

    zznate committed Oct 18, 2010
Commits on Oct 17, 2010
  1. added back sleep btwn host skips, clientPool now keeps track of activ…

    zznate committed Oct 17, 2010
    …e connections
  2. latest round of tweaks after some load testing. retry service fully f…

    zznate committed Oct 17, 2010
    …unctional
Commits on Oct 15, 2010
  1. most test cases now working

    zznate committed Oct 15, 2010
  2. first pass at pushing through new pool structure. Kinda ugly, tests b…

    zznate committed Oct 15, 2010
    …roken mostly, more for proof of concept
Commits on Oct 14, 2010
  1. added RR policy

    zznate committed Oct 14, 2010
  2. least active works

    zznate committed Oct 14, 2010
Commits on Oct 12, 2010
  1. latest round of refactoring for pool plumbing

    zznate committed Oct 12, 2010
Commits on Oct 10, 2010
  1. new client structure, pool isolation

    zznate committed Oct 10, 2010
Commits on Oct 7, 2010
  1. Improve Clock Resolution

    rantav committed Oct 7, 2010
    kudos to jancona and patricioe
Commits on Oct 6, 2010
  1. SlicePredicate: Added support for 'default' ranges

    rantav committed Oct 6, 2010
    When using SlicePredicate for range queries, Hector expects both start and
    finish values to be set. This is not the best behavior.
    For example, Cassandra allows to query for the 'first N columns' simply by
    setting start and finish values to empty string and setting count to N.
    Although this query can be performed with String/StringSerializer, with
    UUID/UUIDSerializer it can't be done. To get 'first N columns' of type UUID,
    one has to generate min/max UUID's and use these as range values.
    
    Cherrypicked from http://github.com/abelaska/hector/commit/48186610cf9275bd4fbd0b14271f94437ce838d4
    Author: tyruk
    
    Note: the actual functionality of default ranges was already enabled, but this commit makes it prettier code-wise
  2. Fixed SuperSliceQuery semantics

    rantav committed Oct 6, 2010
    SuperSliceQuery had wrong semantics. It allowed to query subcolumns slices, not
    super-columns as one would expect.
    
    Cherrypicked from http://github.com/abelaska/hector/commit/fa4e3e3429fc625c40ba7a33d46731a3808308d0
    Work done by tyruk
  3. Validate that HSlicePredicate.count is set before creating a SliceRan…

    rantav committed Oct 6, 2010
    …ge with count of 0.
    
    Thanks B. Todd Burruss for the heads up and suggested solution
  4. added concurrent CCPbH implementation w/ ~ 8% performance improvement…

    zznate committed Oct 6, 2010
    … in stress testing and significantly cleaner dead client and host removal semantics
  5. updated with notes

    zznate committed Oct 6, 2010
  6. updated t version in test

    zznate committed Oct 6, 2010
  7. addressed outofbounds bug from issue 74

    zznate committed Oct 6, 2010
Commits on Oct 5, 2010
  1. added additional index clause arguments

    zznate committed Oct 5, 2010
  2. added additional index clause arguments

    zznate committed Oct 5, 2010
Commits on Oct 2, 2010
  1. updated cassandra jar version

    zznate committed Oct 2, 2010
Commits on Oct 1, 2010