Permalink
Commits on May 8, 2014
  1. DEBUG embstr-classes added.

    antirez committed May 8, 2014
  2. Use a free list for EMBSTR objects creation / release.

    antirez committed May 8, 2014
    To my surprise, mantaining a small poll of objects provides a very big
    speedup even when the jemalloc allocator is used, which is supposed to
    have arenas doing mostly the same work.
    
    Probably this is a combination of different factors:
    
    1) No need to manage zmalloc total memory used.
    2) No locking of any kind because we can assume single thread, so
       maintaining the free list may be cheaper for us compared to
       jemalloc.
    3) More locality because of a different cache/reuse pattern?
Commits on May 7, 2014
  1. REDIS_ENCODING_EMBSTR_SIZE_LIMIT set to 39.

    antirez committed May 7, 2014
    The new value is the limit for the robj + SDS header + string +
    null-term to stay inside the 64 bytes Jemalloc arena in 64 bits
    systems.
  2. Scripting test: check that Lua can call commands rewirting argv.

    antirez committed May 7, 2014
    SPOP, tested in the new test, is among the commands rewritng the
    client->argv argument vector (it gets rewritten as SREM) for command
    replication purposes.
    
    Because of recent optimizations to client->argv caching in the context
    of the Lua internal Redis client, it is important to test for SPOP to be
    callable from Lua without bad effects to the other commands.
  3. Test: handle new osx 'leaks' error.

    antirez committed May 7, 2014
    Sometimes the process is still there but no longer in a state that can
    be checked (after being killed). This used to happen after a call to
    SHUTDOWN NOSAVE in the scripting unit, causing a false positive.
  4. Scripting: objects caching for Lua c->argv creation.

    antirez committed May 6, 2014
    Reusing small objects when possible is a major speedup under certain
    conditions, since it is able to avoid the malloc/free pattern that
    otherwise is performed for every argument in the client command vector.
  5. Scripting: Use faster API for Lua client c->argv creation.

    antirez committed May 6, 2014
    Replace the three calls to Lua API lua_tostring, lua_lua_strlen,
    and lua_isstring, with a single call to lua_tolstring.
    
    ~ 5% consistent speed gain measured.
  6. Scripting: don't call lua_gc() after Lua script run.

    antirez committed May 6, 2014
    Calling lua_gc() after every script execution is too expensive, and
    apparently does not make the execution smoother: the same peak latency
    was measured before and after the commit.
    
    This change accounts for scripts execution speedup in the order of 10%.
  7. Scripting: cache argv in luaRedisGenericCommand().

    antirez committed May 6, 2014
    ~ 4% consistently measured speed improvement.
  8. Fixed missing c->bufpos reset in luaRedisGenericCommand().

    antirez committed May 6, 2014
    Bug introduced when adding a fast path to avoid copying the reply buffer
    for small replies that fit into the client static buffer.
  9. Scripting: replace tolower() with faster code in evalGenericCommand().

    antirez committed May 6, 2014
    The function showed up consuming a non trivial amount of time in the
    profiler output. After this change benchmarking gives a 6% speed
    improvement that can be consistently measured.
  10. Scripting: luaRedisGenericCommand() fast path for buffer-only replies.

    antirez committed May 5, 2014
    When the reply is only contained in the client static output buffer, use
    a fast path avoiding the dynamic allocation of an SDS string to
    concatenate the client reply objects.
  11. Scripting: simpler reply buffer creation in luaRedisGenericCommand().

    antirez committed May 5, 2014
    It if faster to just create the string with a single sdsnewlen() call.
    If c->bufpos is zero, the call will simply be like sdsemtpy().
Commits on May 2, 2014
  1. Cluster: Tcl cluster client: slots-nodes map and close method.

    antirez committed May 2, 2014
    Now the client is able to actually run commands in a Redis Cluster
    assuming the slots->nodes map is stable.
Commits on Apr 30, 2014
Commits on Apr 29, 2014
  1. CLUSTER SET-CONFIG-EPOCH implemented.

    antirez committed Apr 29, 2014
    Initially Redis Cluster accepted that after cluster creation all the
    nodes were at configEpoch 0, evolving from zero as failovers happen.
    
    However later the semantic was made more strict in order to make sure a
    cluster has always all the master nodes with a different configEpoch,
    which is more robust in some corner case (especially resulting from
    errors by the system administrator).
    
    To assign different configEpochs to different nodes at startup was a
    task performed naturally by the config conflicts resolution algorithm
    (see the Cluster specification). However this works well only for small
    clusters or when there are actually just a few collisions, since it is
    designed for exceptional cases.
    
    When a large cluster is created hundred of nodes can be at epoch 0, so
    the conflict resolution code is slow to provide an unique config to each
    node. For this reason this new command was introduced. It can be called
    only when a node is totally fresh: no other nodes known, and configEpoch
    set to zero, so it is safe even against misuses.
    
    redis-trib will use the new command in order to start the cluster
    already setting an incremental unique config to every node.
  2. Cluster test: use 20 instances.

    antirez committed Apr 29, 2014
    This makes tests a bit slower, but it is better to test things at a
    decent scale instead of using just a few nodes, and for a few tests we
    actually need so many nodes.
Commits on Apr 28, 2014
  1. CLIENT LIST speedup via peerid caching + smart allocation.

    antirez committed Apr 28, 2014
    This commit adds peer ID caching in the client structure plus an API
    change and the use of sdsMakeRoomFor() in order to improve the
    reallocation pattern to generate the CLIENT LIST output.
    
    Both the changes account for a very significant speedup.
  2. Added new sdscatfmt() %u and %U format specifiers.

    antirez committed Apr 28, 2014
    This commit also fixes a bug in the implementation of sdscatfmt()
    resulting from stale references to the SDS string header after
    sdsMakeRoomFor() calls.
  3. sdscatfmt() added to SDS library.

    antirez committed Apr 28, 2014
    sdscatprintf() relies on printf() family libc functions and is sometimes
    too slow in critical code paths. sdscatfmt() is an alternative which is:
    
    1) Far less capable.
    2) Format specifier uncompatible.
    3) Faster.
    
    It is suitable to be used in those speed critical code paths such as
    CLIENT LIST output generation.
Commits on Apr 24, 2014