hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yu Li (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-15161) Umbrella: Miscellaneous improvements from production usage
Date Wed, 13 Apr 2016 13:20:25 GMT

    [ https://issues.apache.org/jira/browse/HBASE-15161?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15239215#comment-15239215
] 

Yu Li commented on HBASE-15161:
-------------------------------

[~enis] I tested with your {{CountersBenchmark}} and the default JMH options for 1 and 16
threads, and the result looks a little bit different from yours:
* 1 thread
{noformat}
# JMH 1.12 (released 12 days ago)
# VM version: JDK 1.7.0_45, VM 24.45-b08
# VM invoker: /home/hadoop/hadoop_hbase/jdk1.7.0_45/jre/bin/java
# VM options: <none>
# Warmup: 20 iterations, 1 s each
# Measurement: 20 iterations, 1 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Throughput, ops/time
{noformat}
||Benchmark||Mode||Cnt||Score||Error||Units||
|CountersBenchmark.testAddHistogramConstantValue|thrpt|200|8021489.491|±|28496.232 ops/s|
|CountersBenchmark.testAddHistogramRandomValue|thrpt|200|7264625.900|±|20687.454 ops/s|
|CountersBenchmark.testAddToBlockingQueue|thrpt|200|15193156.700|±|9852.702 ops/s|
|CountersBenchmark.testIncrementAtomicLong|thrpt|200|61286471.569|±|157590.540 ops/s|
|CountersBenchmark.testIncrementCounter|thrpt|200|40704390.815|±|92169.826 ops/s|

* 16 threads:
{noformat}
# JMH 1.12 (released 12 days ago)
# VM version: JDK 1.7.0_45, VM 24.45-b08
# VM invoker: /home/hadoop/hadoop_hbase/jdk1.7.0_45/jre/bin/java
# VM options: <none>
# Warmup: 20 iterations, 1 s each
# Measurement: 20 iterations, 1 s each
# Timeout: 10 min per iteration
# Threads: 16 threads, will synchronize iterations
# Benchmark mode: Throughput, ops/time
{noformat}
||Benchmark||Mode||Cnt||Score||Error||Units||
|CountersBenchmark.testAddHistogramConstantValue|thrpt|200|7943543.369|±|495590.594 ops/s|
|CountersBenchmark.testAddHistogramRandomValue|thrpt|200|5383866.653|±|96536.075 ops/s|
|CountersBenchmark.testAddToBlockingQueue|thrpt|200|6946474.250|±|44876.268 ops/s|
|CountersBenchmark.testIncrementAtomicLong|thrpt|200|12637791.647|±|107859.658 ops/s|
|CountersBenchmark.testIncrementCounter|thrpt|200|605886122.169|±|2663966.723 ops/s|

>From my result it seems MutableHistogram is much less expensive than BlockingQueue, and
performs good on heavy load.

> Umbrella: Miscellaneous improvements from production usage
> ----------------------------------------------------------
>
>                 Key: HBASE-15161
>                 URL: https://issues.apache.org/jira/browse/HBASE-15161
>             Project: HBase
>          Issue Type: Umbrella
>            Reporter: Yu Li
>            Assignee: Yu Li
>
> We use HBase to (mainly) build index for our search engine in Alibaba. Recently we are
upgrading our online cluster from 0.98.12 to 1.x and I'd like to take the opportunity to contribute
a bunch of our private patches to community (better late than never, I hope :-)). This is
an umbrella to track this effort.



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

Mime
View raw message