[ 
https://issues.apache.org/jira/browse/IGNITE-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15082693#comment-15082693
 ] 

Vladimir Ozerov commented on IGNITE-2333:
-----------------------------------------

Quick and dirty prototype. Shown ~10-15% increase in throughput for single-node 
PUT benchmark.

{code}
public class StripedRWL implements ReadWriteLock {

    private final PaddedRWL[] locks;

    public StripedRWL() {
        locks = new PaddedRWL[16];

        for (int i = 0; i < locks.length; i++)
            locks[i] = new PaddedRWL();
    }

    private static final ThreadLocal<Integer> IDX = new ThreadLocal<Integer>() {
        @Override protected Integer initialValue() {
            return ThreadLocalRandom.current().nextInt(100000) % 16;
        }
    };

    @NotNull @Override public Lock readLock() {
        return locks[IDX.get()].readLock();
    }

    @NotNull @Override public Lock writeLock() {
        return locks[IDX.get()].writeLock();
    }

    private static class PaddedRWL extends ReentrantReadWriteLock {
        long p0, p1, p2, p3, p4, p5, p6, p7;
    }
}
{code}

> Hotspot in GridDhtPartitionTopologyImpl readLock() and readUnlock() methods
> ---------------------------------------------------------------------------
>
>                 Key: IGNITE-2333
>                 URL: https://issues.apache.org/jira/browse/IGNITE-2333
>             Project: Ignite
>          Issue Type: Sub-task
>          Components: cache, general
>    Affects Versions: ignite-1.4
>            Reporter: Vladimir Ozerov
>             Fix For: 1.6
>
>
> When running simple PUT benchmark in ATOMIC cache, read lock-unlock consumes 
> up to 10% of time.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to