[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14720350#comment-14720350 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-0.98 #1103 (See [https://builds.apache.org/job/HBase-0.98/1103/]) HBASE-14289 Backport HBASE-13965 'Stochastic Load Balancer JMX Metrics' to 0.98 (tedyu: rev 7a0d36fecfebb1f61811181347f90a9b88a9d09b) * hbase-hadoop1-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-hadoop1-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManager.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: 13965-addendum.txt, HBASE-13965-branch-1-v2.patch, HBASE-13965-branch-1.patch, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14720429#comment-14720429 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-0.98-on-Hadoop-1.1 #1057 (See [https://builds.apache.org/job/HBase-0.98-on-Hadoop-1.1/1057/]) HBASE-14289 Backport HBASE-13965 'Stochastic Load Balancer JMX Metrics' to 0.98 (tedyu: rev 7a0d36fecfebb1f61811181347f90a9b88a9d09b) * hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManager.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-hadoop1-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-hadoop1-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: 13965-addendum.txt, HBASE-13965-branch-1-v2.patch, HBASE-13965-branch-1.patch, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14659522#comment-14659522 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-1.3 #90 (See [https://builds.apache.org/job/HBase-1.3/90/]) HBASE-13965 Stochastic Load Balancer JMX Metrics (Lei Chen) (tedyu: rev 6a2b618d97c13320bb91a5f555827437099e3a26) * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManager.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-branch-1-v2.patch, HBASE-13965-branch-1.patch, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14659463#comment-14659463
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12748968/HBASE-13965-branch-1-v2.patch
against branch-1 branch at commit 18c9bb8b54a5bb101f793e555272cbc74b8288cb.
ATTACHMENT ID: 12748968
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 12 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
org.apache.hadoop.hbase.util.TestProcessBasedCluster
org.apache.hadoop.hbase.mapreduce.TestImportExport
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14987//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14987//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14987//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14987//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Fix For: 2.0.0
Attachments: 13965-addendum.txt, HBASE-13965-branch-1-v2.patch,
HBASE-13965-branch-1.patch, HBASE-13965-v10.patch, HBASE-13965-v11.patch,
HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch,
HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch,
HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14659508#comment-14659508 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-1.3-IT #72 (See [https://builds.apache.org/job/HBase-1.3-IT/72/]) HBASE-13965 Stochastic Load Balancer JMX Metrics (Lei Chen) (tedyu: rev 6a2b618d97c13320bb91a5f555827437099e3a26) * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManager.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-branch-1-v2.patch, HBASE-13965-branch-1.patch, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14659136#comment-14659136
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12748903/HBASE-13965-branch-1.patch
against branch-1 branch at commit 18c9bb8b54a5bb101f793e555272cbc74b8288cb.
ATTACHMENT ID: 12748903
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 12 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 lineLengths{color}. The patch introduces the following lines
longer than 100:
+ for (EntryTableName, MapServerName, ListHRegionInfo e :
assignmentsByTable.entrySet()) {
+
JMXConnectorFactory.connect(JMXListener.buildJMXServiceURL(connectorPort,
connectorPort));
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14984//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14984//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14984//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14984//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Fix For: 2.0.0
Attachments: 13965-addendum.txt, HBASE-13965-branch-1.patch,
HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch,
HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch,
HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch,
HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652037#comment-14652037 ] Ted Yu commented on HBASE-13965: Planning to commit once QA run passes. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: HBASE-13965-v10.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652688#comment-14652688 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-TRUNK #6694 (See [https://builds.apache.org/job/HBase-TRUNK/6694/]) HBASE-13965 Stochastic Load Balancer JMX Metrics (Lei Chen) (tedyu: rev 20d1fa36e7ffa1c8d274def831223bff9b04fa69) * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java HBASE-13965 Stochastic Load Balancer JMX Metrics (Lei Chen) (tedyu: rev 598cfeb77563a3fea9d0ed467025514662e52ca0) * hbase-client/src/main/java/org/apache/hadoop/hbase/filter/Filter.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-client/src/main/java/org/apache/hadoop/hbase/filter/FilterWrapper.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/ScanQueryMatcher.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652765#comment-14652765 ] Lei Chen commented on HBASE-13965: -- +1 Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652936#comment-14652936 ] Hudson commented on HBASE-13965: SUCCESS: Integrated in HBase-1.3-IT #68 (See [https://builds.apache.org/job/HBase-1.3-IT/68/]) HBASE-13965 Revert due to test failure in TestAssignmentManager (tedyu: rev 24dbe25e95d0a355b2e07aa94b5921ff4b4865e9) * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652658#comment-14652658
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12748522/HBASE-13965-v11.patch
against master branch at commit 4b6598e394bae67b54d6f741dd262afe03b2c133.
ATTACHMENT ID: 12748522
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 8 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14964//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14964//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14964//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14964//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Fix For: 2.0.0, 1.3.0
Attachments: HBASE-13965-v10.patch, HBASE-13965-v11.patch,
HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch,
HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch,
HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652947#comment-14652947 ] Hudson commented on HBASE-13965: SUCCESS: Integrated in HBase-1.3 #86 (See [https://builds.apache.org/job/HBase-1.3/86/]) HBASE-13965 Revert due to test failure in TestAssignmentManager (tedyu: rev 24dbe25e95d0a355b2e07aa94b5921ff4b4865e9) * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652654#comment-14652654 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-1.3 #85 (See [https://builds.apache.org/job/HBase-1.3/85/]) HBASE-13965 Stochastic Load Balancer JMX Metrics (Lei Chen) (tedyu: rev c215b900f49685989083df3786bd8441700c248a) * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652650#comment-14652650 ] Hudson commented on HBASE-13965: SUCCESS: Integrated in HBase-1.3-IT #67 (See [https://builds.apache.org/job/HBase-1.3-IT/67/]) HBASE-13965 Stochastic Load Balancer JMX Metrics (Lei Chen) (tedyu: rev c215b900f49685989083df3786bd8441700c248a) * hbase-server/src/test/java/org/apache/hadoop/hbase/master/balancer/TestBaseLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/FavoredNodeLoadBalancer.java * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancer.java * hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSource.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/LoadBalancer.java * hbase-hadoop2-compat/src/main/resources/META-INF/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/StochasticLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/SimpleLoadBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsBalancer.java * hbase-server/src/main/java/org/apache/hadoop/hbase/master/balancer/BaseLoadBalancer.java * hbase-common/src/main/java/org/apache/hadoop/hbase/HConstants.java * hbase-hadoop2-compat/src/main/java/org/apache/hadoop/hbase/master/balancer/MetricsStochasticBalancerSourceImpl.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652720#comment-14652720 ] Ted Yu commented on HBASE-13965: Addendum deals with the scenario where connector port has been taken on the test machine. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14653005#comment-14653005 ] Hudson commented on HBASE-13965: FAILURE: Integrated in HBase-TRUNK #6695 (See [https://builds.apache.org/job/HBase-TRUNK/6695/]) HBASE-13965 Addendum tries different connector ports if BindException is encountered (tedyu: rev 931e77d4507e1650c452cefadda450e0bf3f0528) * hbase-server/src/test/java/org/apache/hadoop/hbase/TestStochasticBalancerJmxMetrics.java Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0 Attachments: 13965-addendum.txt, HBASE-13965-v10.patch, HBASE-13965-v11.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652252#comment-14652252
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12748476/HBASE-13965-v10.patch
against master branch at commit 4b6598e394bae67b54d6f741dd262afe03b2c133.
ATTACHMENT ID: 12748476
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 8 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:red}-1 release audit{color}. The applied patch generated 1 release
audit warnings (more than the master's current 0 warnings).
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
{color:red}-1 core zombie tests{color}. There are 1 zombie test(s):
at
org.apache.sentry.tests.e2e.hive.hiveserver.AbstractHiveServer.createConnection(AbstractHiveServer.java:63)
at
org.apache.sentry.tests.e2e.hive.Context.createConnection(Context.java:92)
at
org.apache.sentry.tests.e2e.hive.AbstractTestWithStaticConfiguration.setupAdmin(AbstractTestWithStaticConfiguration.java:472)
at
org.apache.sentry.tests.e2e.dbprovider.TestDatabaseProvider.testGrantRevokeRoleToGroups(TestDatabaseProvider.java:2037)
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14963//testReport/
Release audit warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14963//artifact/patchprocess/patchReleaseAuditWarnings.txt
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14963//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14963//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14963//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Fix For: 2.0.0, 1.3.0
Attachments: HBASE-13965-v10.patch, HBASE-13965-v3.patch,
HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch,
HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch,
HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652314#comment-14652314 ] Lei Chen commented on HBASE-13965: -- thanks, I will update soon Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: HBASE-13965-v10.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14652270#comment-14652270 ] Ted Yu commented on HBASE-13965: @Lei : The following file triggered release audit warning: hbase-hadoop2-compat/src/main/resources/x/services/org.apache.hadoop.hbase.master.balancer.MetricsStochasticBalancerSource Take a look at ./hbase-hadoop2-compat/src/test/resources/META-INF/services/org.apache.hadoop.hbase.HadoopShims to see how license is added. I will commit the next patch with the above fix. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Fix For: 2.0.0, 1.3.0 Attachments: HBASE-13965-v10.patch, HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14643092#comment-14643092 ] Ted Yu commented on HBASE-13965: StochasticLoadBalancer related tests passed. The two failed tests are not related to the patch. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14643084#comment-14643084
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12747339/HBASE-13965-v9.patch
against master branch at commit 3f80e0ea4f101298d9dbd574ffcf388afb621d55.
ATTACHMENT ID: 12747339
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 8 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14897//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14897//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14897//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14897//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch,
HBase-13965-JConsole.png, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14643544#comment-14643544
]
Ted Yu commented on HBASE-13965:
{code}
+ * This constructor that accepts an instance of MetricsBalancer,
+ * which will be used instead of creating a new one
{code}
'This constructor' - 'Constructor'
{code}
+ protected BaseLoadBalancer(MetricsBalancer metricsBalancer) {
+if (metricsBalancer != null)
+ this.metricsBalancer = metricsBalancer;
+else
+ this.metricsBalancer = new MetricsBalancer();
{code}
Either put the assignment on the same line as if / else, or enclose assignment
in curly braces.
{code}
+} catch (Exception e) {
+ LOG.info(failed to get the size of all tables, exception = +
e.getMessage());
{code}
The above log should be at error level.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965-v8.patch, HBASE-13965-v9.patch, HBASE-13965_v2.patch,
HBase-13965-JConsole.png, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14641992#comment-14641992 ] Ted Yu commented on HBASE-13965: https://builds.apache.org/job/PreCommit-HBASE-Build/14771//testReport is no longer accessible. Mind attaching patch again ? Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14642069#comment-14642069 ] Lei Chen commented on HBASE-13965: -- I will attach a summary and a patch soon. Thanks for reminding. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-JConsole.png, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14639877#comment-14639877 ] Lei Chen commented on HBASE-13965: -- I've been trying to apply the patch on a test hbase cluster for verifying. Currently experiencing some compatibility issues. Will update status once got results. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14636112#comment-14636112 ] Enis Soztutar commented on HBASE-13965: --- bq. Lets just fully remove the per table balancing. I've wanted to do that for a while. Sounds good. Lets do a follow jira after this. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14627088#comment-14627088 ] Elliott Clark commented on HBASE-13965: --- bq.Do you mean removing all the code for the per-table balancing, as well as documents if any? Yep, that greatly simplifies lots of things, and cleans up some code that was put in as a stop gap measure. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14627012#comment-14627012 ] Ted Yu commented on HBASE-13965: w.r.t. test failure of TestWALProcedureStoreOnHDFS: it is not related to your patch. For #3, it should be done in separate JIRA. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14627240#comment-14627240
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12745330/HBASE-13965-v8.patch
against master branch at commit 2f327c911056d02813f642503db9a4383e8b4a2f.
ATTACHMENT ID: 12745330
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 8 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
{color:red}-1 core zombie tests{color}. There are 1 zombie test(s):
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14771//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14771//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14771//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14771//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965-v8.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14624668#comment-14624668 ] Lei Chen commented on HBASE-13965: -- Do you mean removing all the code for the per-table balancing, as well as documents if any? If the balancing is always performed on the ensemble table, then no table name is needed for the metrics name in JMX, right? Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14623975#comment-14623975
]
Ted Yu commented on HBASE-13965:
Including costs for system table in JMX metrics is fine.
{code}
+if (!t.getTableName().isSystemTable()) {
+ tablesCount++;
+}
{code}
In that case the above can be removed.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14624048#comment-14624048
]
Lei Chen commented on HBASE-13965:
--
Thanks for pointing it out.
I had some test and found that there are two system tables, hbase:meta and
hbase:namespace.
The first one, hbase:meta, will not be given to balancer while hbase:namespace
will.
Maybe I can just get the number of all tables, and -1 for removing hbase:meta?
That will avoid the looping.
{code}tablesCount = services.getTableDescriptors().getAll().size();
// -1 for removing a system table, hbase:meta, which will not be balanced.
tablesCount--;
{code}
However, this is based on an assumption that only one non-balancing system
table exists. Is there any better solution?
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14624235#comment-14624235 ] Elliott Clark commented on HBASE-13965: --- Lets just fully remove the per table balancing. I've wanted to do that for a while. bq.However, this is based on an assumption that only one non-balancing system table exists. If we move the security tables currently, then we should stop. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14623933#comment-14623933
]
Lei Chen commented on HBASE-13965:
--
Thanks for the feedback.
{Should we do early return if the table is system table ?}
Should filtering tables happen before/outside updateStochasticCosts? (e.g in
HMaster?)
Since HMaster sends hbase:namespace for balancing, there might be a reason for
that, and I feel it may help to include its costs in JMX, just in case.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14623634#comment-14623634
]
Ted Yu commented on HBASE-13965:
{code}
+ /** Config for balancing the cluster by each table */
{code}
'by each table' - 'by table'
{code}
+ /** The name of the combined table when bytable is off*/
{code}
'combined table' - 'combined tables'
{code}
+ // Use Most Recent Used(MRU) cache
{code}
'Most Recent' - 'Most Recently'
{code}
+synchronized (stochasticCosts) {
+ if (stochasticCosts != null) {
{code}
Looks like the null check is not needed.
{code}
+ private void updateStochasticCosts(TableName tableName, Double overall,
Double[] subCosts) {
{code}
Should we do early return if the table is system table ?
{code}
+ //TODO: should hbase:mespace be passed in by HMaster?
{code}
Can the above TODO be dropped ?
{code}
+ ((MetricsStochasticBalancer)
metricsBalancer).updateStochasticCost(tableName.getNameAsString(), Overall,
+Overall cost, overall);
{code}
Since metricsBalancer would be cast in the loop that follows, you can use local
variable of type MetricsStochasticBalancer to save the casts in the loop.
{code}
+if (null == tableName)
+ return;
{code}
nit: move return on the same line as if statement to save a pair of curly
braces.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14623151#comment-14623151
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12744790/HBASE-13965-v7.patch
against master branch at commit c16bbf47cbb1017b92960e15edfaa81cfd104b1d.
ATTACHMENT ID: 12744790
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 8 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 lineLengths{color}. The patch introduces the following lines
longer than 100:
+ if
(!server.getConfiguration().getBoolean(HConstants.HBASE_MASTER_LOADBALANCE_BYTABLE,
false)) {
+ ((MetricsStochasticBalancer)
metricsBalancer).updateStochasticCost(tableName.getNameAsString(), Overall,
+((MetricsStochasticBalancer)
metricsBalancer).updateStochasticCost(tableName.getNameAsString(),
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
org.apache.hadoop.hbase.master.procedure.TestWALProcedureStoreOnHDFS
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14745//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14745//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14745//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14745//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622376#comment-14622376 ] Ted Yu commented on HBASE-13965: w.r.t. ensemble mode, one option is to leave table name blank. We then have: _Overall _costFunction X N Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622624#comment-14622624 ] stack commented on HBASE-13965: --- If a table is being balanced by the balancer, I would think an operator would want to know metrics on why. A blank table name given table is not a dimension when in ensemble mode would be reasonable -- I'd remove the '_' prefix though... in fact, so we don't surprise operators, what about flipping the metric name so it is Overall_Table1 rather than Table1_Overall. If we start with Overall, operator will know where to look for this metric in the list... if table prefix, could be anywhere in the list? Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622705#comment-14622705 ] Lei Chen commented on HBASE-13965: -- Ted, Clay, and stack, thanks for offering suggestions. Currently, in the ensemble mode, the combined table is named ensemble, I somehow feel that it might be a good idea to use the name directly. For example, ensemble_Overall or ensemble_costFunction1. In a similar way, when in the per-table mode, the system table hbase:namespace can be reported to JMX using it's name directly as well. The benefit is that we can avoid any special case handling in the balancer. The balancer just calculate and report whatever tables given by hmaster. I'm updating the unit test to cover both modes. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622773#comment-14622773 ] stack commented on HBASE-13965: --- Is ensemble a system table? Is it a logical table only? Prefix with system namespace? Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622782#comment-14622782 ] Lei Chen commented on HBASE-13965: -- I think the ensemble table is more like a temporary table, and it is not in any namespace. https://github.com/apache/hbase/blob/0.98/hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java#L792 Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622894#comment-14622894 ] stack commented on HBASE-13965: --- bq. I think the ensemble table is more like a temporary table, and it is not in any namespace. It is not a table at all -- smile. It is a key in a map usually keyed by tablename... it can be anything it looks like. So, put it in system namespace and then it won't clash with anyone or put it in an impossible namespace, 'ephemeral'. See here for what is legal: http://grepcode.com/file/repo1.maven.org/maven2/org.apache.hbase/hbase-common/1.0.0/org/apache/hadoop/hbase/TableName.java#70 Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14622900#comment-14622900 ] Lei Chen commented on HBASE-13965: -- Thanks for clarifying. I will use hbase:ensemble for this cute table-like thing - smile. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965-v7.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14621657#comment-14621657 ] Ted Yu commented on HBASE-13965: I assume user(s) would not be interested in the costs for system tables. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14621641#comment-14621641
]
Lei Chen commented on HBASE-13965:
--
I also found that when {{hbase.master.loadbalancer.bytable}} is set to true,
balancing will also be performed on table hbase:namespace, which is a system
table. Should the costs of hbase:namespace be reported to JMX the same way as
user tables?
Any idea?
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14621457#comment-14621457
]
Lei Chen commented on HBASE-13965:
--
I have found a problem related to HBASE-5231(per-table load balancing).
It seems that the balancing is done by an iteration of tables.
https://github.com/apache/hbase/blob/master/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java#L1219-L1228
This can be configured to be in per-table mode or ensemble mode.
https://github.com/apache/hbase/blob/master/hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java#L956-L962
In ensemble mode, all the tables are copied into an ensemble table for
balancing. The configuration is set by
{{hbase.master.loadbalancer.bytable}}
My question is how to name the metrics if the balancing is in ensemble mode.
For example, suppose we have two tables, Table1 and Table2, and N cost
functions.
In the per-table mode, each table will have an overall cost and one for each
cost function.
{{Table1_Overall}}
{{Table1_costFunction}} x N
{{Table2_Overall}}
{{Table2_costFunction}} x N
In the ensemble mode, there will be only one overall and one set of function
costs.
{{ensemble_Overall}}
{{ensemble_costFunction}} x N
Can we use a special name for the combined table, e.g. ensemble? The problem
is that the user may have already created a table named ensemble, which may
cause confusion.
Any idea on this problem?
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14618960#comment-14618960
]
Ted Yu commented on HBASE-13965:
{code}
+ if (tableName != null) break;
+}
+ if (tableName != null) break;
{code}
You can return tableName on line of the first break. This way the second return
won't be needed.
For CostFunctions whose value is 0, can we omit them in the metrics ?
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14618991#comment-14618991
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12744232/HBASE-13965-v6.patch
against master branch at commit f5ad736282c8c9c27b14131919d60b72834ec9e4.
ATTACHMENT ID: 12744232
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:red}-1 core tests{color}. The patch failed these unit tests:
org.apache.hadoop.hbase.TestJMXListener
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14701//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14701//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14701//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14701//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14618997#comment-14618997 ] Lei Chen commented on HBASE-13965: -- Yes, sounds good, since the full list of cost functions should be known to the user. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14619044#comment-14619044 ] stack commented on HBASE-13965: --- bq. For CostFunctions whose value is 0, can we omit them in the metrics ? So metrics can come and go (not show if zero then show when non-zero)? That will be confusing I'd say. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14619064#comment-14619064 ] Clay B. commented on HBASE-13965: - [~tedyu] I agree with [~stack] as for Graphite there which I load my metrics into, there is a difference between a null (no value) and a zero. Often we run arithmetic functions on our metrics for alarming for example. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14619554#comment-14619554 ] Lei Chen commented on HBASE-13965: -- Thanks [~clayb] for giving suggestion. I have found that the stochastic load balancer holds a reference to HMaster, which can be used to get the number of tables, therefore the size of the map can be determined. No need to use configurable value. I will update the patch soon. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14617464#comment-14617464
]
Ted Yu commented on HBASE-13965:
Loaded patch v5 on a small cluster and obtained the following:
{code}
}, {
name : Hadoop:service=HBase,name=Master,sub=Balancer,
modelerType : Master,sub=Balancer,
tag.Context : master,
tag.Hostname : cn013.l42scl.hortonworks.com,
IntegrationTestBigLinkedList_StoreFileCostFunction : 3.262317387568032,
IntegrationTestBigLinkedList_LocalityCostFunction : 2.4739584,
IntegrationTestBigLinkedList_TableSkewCostFunction : 5.60546874999,
IntegrationTestBigLinkedList_Overall : 35.09174447090137,
IntegrationTestBigLinkedList_WriteRequestCostFunction : 0.0,
IntegrationTestBigLinkedList_RegionCountSkewCostFunction : 0.0,
IntegrationTestBigLinkedList_ReadRequestCostFunction : 5.0,
IntegrationTestBigLinkedList_MemstoreSizeCostFunction : 0.0,
IntegrationTestBigLinkedList_RegionReplicaHostCostFunction : 0.0,
IntegrationTestBigLinkedList_RegionReplicaRackCostFunction : 0.0,
IntegrationTestBigLinkedList_MoveCostFunction : 18.75,
{code}
Do you think it makes sense to expose each cost (other than Overall) as
percentage ?
This way, it is easier for user to figure out which cost is the dominant factor.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14617922#comment-14617922 ] Lei Chen commented on HBASE-13965: -- Thanks for testing the patch and posting the result metrics. I agree that using percentage is easier for quick look. I will update the patch. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14617931#comment-14617931
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12744125/HBASE-13965-v6.patch
against master branch at commit f5ad736282c8c9c27b14131919d60b72834ec9e4.
ATTACHMENT ID: 12744125
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified tests.
{color:red}-1 patch{color}. The patch command could not apply the patch.
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14697//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965-v6.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14615265#comment-14615265
]
Ted Yu commented on HBASE-13965:
{code}
32public void updateStochasticCost(String tableName, String
costFunctionName,
{code}
Should tableName parameter be of type TableName ?
For MetricsStochasticBalancerSourceImpl.java , it should be annotated with
@InterfaceAudience.Private
{code}
56 * The function that report stochastic load balancer costs to JMX
{code}
'that report' - 'that reports'
{code}
58public void updateStochasticCost(String tableName, String
costFunctionName,
59String costFunctionDesc, Double value) {
{code}
costFunctionDesc isn't used in the method.
{code}
82String attrName = tableName + ((tableName.length() = 0) ?
: TABLE_FUNCTION_SEP) + key;
{code}
When would tableName.length() be = 0 ? There is check at the beginning of
updateStochasticCost().
{code}
124 private Double[] lastSubcosts;
{code}
nit: Uppercase 'c' of 'costs'
{code}
470 this.lastSubcosts[i] = multiplier*cost;
471 total += multiplier * cost;
{code}
There is no need to perform the same multiplication twice.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14615518#comment-14615518 ] Lei Chen commented on HBASE-13965: -- Thanks for your review and great feedback. I will update an updated patch. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14615467#comment-14615467
]
Hadoop QA commented on HBASE-13965:
---
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12743741/HBASE-13965-v3.patch
against master branch at commit cff1a5f1f5cc8f5e1e99c6aecb39e2e69f86bf7e.
ATTACHMENT ID: 12743741
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:green}+1 core tests{color}. The patch passed unit tests in .
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14678//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14678//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14678//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14678//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965_v2.patch,
HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14615805#comment-14615805
]
Ted Yu commented on HBASE-13965:
{code}
79 costs.put(costFunctionName, descAndValue);
80 stochasticCosts.put(tableName, costs);
{code}
For any specific cost function, its description should be fixed.
In the above model, description for the same cost function is stored once per
table.
Can we reduce the number of times description for same function is stored ?
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14615800#comment-14615800
]
Hadoop QA commented on HBASE-13965:
---
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12743774/HBASE-13965-v4.patch
against master branch at commit 608c3aa15c34b9014f99e857b374645db58cbbe3.
ATTACHMENT ID: 12743774
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:green}+1 core tests{color}. The patch passed unit tests in .
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14680//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14680//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14680//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14680//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14615909#comment-14615909 ] Lei Chen commented on HBASE-13965: -- Good point. It can be more memory efficient if description is stored only once for each cost function. Patch will be updated. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14616118#comment-14616118
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12743848/HBASE-13965-v5.patch
against master branch at commit c220635c7893c96db675cb2b80af6ade4a44e3d4.
ATTACHMENT ID: 12743848
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 checkstyle{color}. The applied patch does not increase the
total number of checkstyle errors
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:green}+1 core tests{color}. The patch passed unit tests in .
{color:red}-1 core zombie tests{color}. There are 1 zombie test(s):
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14685//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14685//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14685//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14685//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965-v3.patch, HBASE-13965-v4.patch,
HBASE-13965-v5.patch, HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14614448#comment-14614448
]
Hadoop QA commented on HBASE-13965:
---
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12743132/HBASE-13965_v2.patch
against master branch at commit 1f9bf419c1c9040504de1b36fb85d0be7163eadf.
ATTACHMENT ID: 12743132
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified tests.
{color:green}+1 hadoop versions{color}. The patch compiles with all
supported hadoop versions (2.4.0 2.4.1 2.5.0 2.5.1 2.5.2 2.6.0 2.7.0)
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 protoc{color}. The applied patch does not increase the
total number of protoc compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:red}-1 checkstyle{color}. The applied patch generated
1903 checkstyle errors (more than the master's current 1898 errors).
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 lineLengths{color}. The patch does not introduce lines
longer than 100
{color:green}+1 site{color}. The mvn post-site goal succeeds with this patch.
{color:green}+1 core tests{color}. The patch passed unit tests in .
Test results:
https://builds.apache.org/job/PreCommit-HBASE-Build/14670//testReport/
Release Findbugs (version 2.0.3)warnings:
https://builds.apache.org/job/PreCommit-HBASE-Build/14670//artifact/patchprocess/newFindbugsWarnings.html
Checkstyle Errors:
https://builds.apache.org/job/PreCommit-HBASE-Build/14670//artifact/patchprocess/checkstyle-aggregate.html
Console output:
https://builds.apache.org/job/PreCommit-HBASE-Build/14670//console
This message is automatically generated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBASE-13965_v2.patch, HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14610601#comment-14610601 ] Lei Chen commented on HBASE-13965: -- I agree that the unused balancers should be purged or made into attributes of the stochastic load balancer. I think it may be better to do it in another Jira, since “One thing at a time”. About the ever growing map, I’m thinking of two ways to solve this problem. 1. Besides updateStochasticCost, add another method (or add a boolean parameter) which should be called when the table is deleted. This will allow the map to contain only existing tables 2. Use a fixed-size most recent used (MRU) cache to store the map. The size can be configurable. Any suggestion? Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14609412#comment-14609412 ] Lei Chen commented on HBASE-13965: -- Thanks for the feedback. I appreciate your detailed review and help. I will modify the patch. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14609392#comment-14609392 ] Enis Soztutar commented on HBASE-13965: --- bq. Feel free to ignore the above opinion. Just change patch so DEFAULT is STOCHASTIC. Then I think that part is gone already, instead the StochasticLB instantiates the object and send it via constructor to the super class. The attached patch is a set of stacked patches. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14609390#comment-14609390
]
Enis Soztutar commented on HBASE-13965:
---
Thanks Lei for the patch. Looks good overall. A couple of feedback items:
- Can you create the patch so that it is a one commit. Current stacked patch
is not reviewable. See https://hbase.apache.org/book.html#submitting.patches
- As per Stack, please only include changes needed for the patch. {{conf/}}
and some others are not needed.
- We usually use: {{(name != null)}} rather than {{(null != name)}}
- This map is ever growing:
{code}
+ MapString, MapString, Double stochasticCosts =
{code}
In a dynamic cluster where tables are short lived, it can cause OOM. Not sure
whether there is an easy way to phase out stuff from that map.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Attachments: HBase-13965-v1.patch,
stochasticloadbalancerclasses_v2.png
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14609072#comment-14609072 ] Clay B. commented on HBASE-13965: - Hi [~stack] as to the DEFAULT versus STOCHASTIC, you're in favor of collapsing the metrics provider to be stochastic-centric opposed to pluggable? If so, I'm a bit concerned as HBASE-6721 proposes a new balancer who's functionality we appreciate and which would likely need some visibility too. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14609007#comment-14609007 ] stack commented on HBASE-13965: --- Thank you for the design and patch overview. Helps. The changes in your patch to hbase-env.sh are probably not intended. Ditto hbase-site.xml... and others. Add license to new classes (see adjacent classes for how), javadoc on Interfaces especially (no need of public qualifiers when an Interface), and see annotations in accompanying classes and apply as appropriate here (mark them private I'd say). I wouldn't even bother offering DEFAULT vs STOCHASTIC. No one runs any balancer but stochastic (Besides default is the Stochastic). Remove the choice. Otherwise, patch looking good. Suggest you past example of what it looks like when it is working. FYI [~eclark] Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14609187#comment-14609187 ] stack commented on HBASE-13965: --- bq. ...you're in favor of collapsing the metrics provider to be stochastic-centric opposed to pluggable? Yes sir. It is what is run by all. We should purge all others (FavoredNodes, GroupBased...) and redo these unused (or rarely used) exotica instead as attributes on the stochastic balancer, the balancer we near-all run (and trust). Feel free to ignore the above opinion. Just change patch so DEFAULT is STOCHASTIC. Then. Thanks [~clayb] Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Attachments: HBase-13965-v1.patch, stochasticloadbalancerclasses_v2.png Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[
https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14606864#comment-14606864
]
Lei Chen commented on HBASE-13965:
--
Sorry for the delay. I propose the following changes which can be seen
illustrated in the attached class diagram.
The {{StochasticLoadBalancer}} extends {{BaseLoadBalancer}} which has a class
variable of type {{MetricsBalancer}}. The {{MetricsBalancer}} contains a
private class variable of type {{MetricsBalancerSource}}. The
{{MetricsBalancerSource}} is an interface which defines which metrics can be
reported to JMX. This proves to make extension difficult for load balancer
implementation specific metrics (e.g. the {{StochasticLoadBalancer}}).
Adding metrics to the generic interface is not appropriate being it is used by
all load balancers and should not contain any load balancer specific metrics. I
propose to create a class extending {{MetricsBalancer}} to provide specific
load balancer metrics. To use this class, I propose to add a constructor to
{{BaseLoadBalancer}} which allows for the balancer instance metrics class to be
passed in. (Thanks [~enis] for code review and giving the constructor
suggestion!)
In the constructor of {{StochasticLoadBalancer}}, an instance of
{{MetricsStochasticBalancer}} is created and passed to a new constructor added
to {{BaseLoadBalancer}}, which will use it to replace the default
{{MetricsBalancer}}. The function used to add metrics is declared as following:
{code}
public void updateStochasticCost(String tableName, String costFunctionName,
String costFunctionDesc, Double value);
{code}
In {{MetricsBalancer}}, the {{private final}} class variable {{source}} was
previously hardcoded and instantiated in its constructor; I propose a new
function {{initSource}} which can be overridden to set this variable. As such,
in the subclass {{MetricsStochasticBalancer}}, {{initSource}} will create a
{{MetricsStochasticBalancerSource}} instance instead of the default
{{MetricsBalancerSource}}.
Finally, to give good insight to the internal status of
{{StochasticLoadBalancer}}, we are considering adding metrics for each cost
function, as well as the overall cost. For example, if the balancing is carried
out for table MyTable1 and the {{StochasticLoadBalancer}} has 3 cost
functions MoveCost, LocalityCost, and RegionReplicaHostCost, then 4
metrics will be added to “HBase - Master - Balancer” as following:
MyTable1_Overall
MyTable1_MoveCost
MyTable1_LocalityCost
MyTable1_RegionReplicaHostCost
I'm building the patch, any suggestion is appreciated.
Stochastic Load Balancer JMX Metrics
Key: HBASE-13965
URL: https://issues.apache.org/jira/browse/HBASE-13965
Project: HBase
Issue Type: Improvement
Components: Balancer, metrics
Reporter: Lei Chen
Assignee: Lei Chen
Today’s default HBase load balancer (the Stochastic load balancer) is cost
function based. The cost function weights are tunable but no visibility into
those cost function results is directly provided.
A driving example is a cluster we have been tuning which has skewed rack size
(one rack has half the nodes of the other few racks). We are tuning the
cluster for uniform response time from all region servers with the ability to
tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and
RegionCountSkew Cost is difficult without a way to attribute each cost
function’s contribution to overall cost.
What this jira proposes is to provide visibility via JMX into each cost
function of the stochastic load balancer, as well as the overall cost of the
balancing plan.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14603804#comment-14603804 ] Andrew Purtell commented on HBASE-13965: Are you planning on providing a patch [~leochen4891]? Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-13965) Stochastic Load Balancer JMX Metrics
[ https://issues.apache.org/jira/browse/HBASE-13965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14600285#comment-14600285 ] Ted Yu commented on HBASE-13965: This is very useful feature. Stochastic Load Balancer JMX Metrics Key: HBASE-13965 URL: https://issues.apache.org/jira/browse/HBASE-13965 Project: HBase Issue Type: Improvement Components: Balancer, metrics Reporter: Lei Chen Assignee: Lei Chen Today’s default HBase load balancer (the Stochastic load balancer) is cost function based. The cost function weights are tunable but no visibility into those cost function results is directly provided. A driving example is a cluster we have been tuning which has skewed rack size (one rack has half the nodes of the other few racks). We are tuning the cluster for uniform response time from all region servers with the ability to tolerate a rack failure. Balancing LocalityCost, RegionReplicaRack Cost and RegionCountSkew Cost is difficult without a way to attribute each cost function’s contribution to overall cost. What this jira proposes is to provide visibility via JMX into each cost function of the stochastic load balancer, as well as the overall cost of the balancing plan. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
