[
https://issues.apache.org/jira/browse/AMBARI-6485?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Azwaw OUSADOU updated AMBARI-6485:
----------------------------------
Attachment: TRUNK-OUTPUT-TEST.txt
AMBARI-6485-OUTPUT-TEST.txt
AMBARI-6485.diff
BUILD and install on centos 6 :
mvn versions:set -DnewVersion=${AMBARI_VERSION}
mvn -B clean install package rpm:rpm -DnewVersion=${AMBARI_VERSION} -DskipTests
-Dpython.ver="python >= 2.6" -Preplaceurl
on ambari-server :
copy
ambari-server/target/rpm/ambari-server/RPMS/noarch/ambari-server-*.noarch.rpm
on server
with root access right : yum install ambari-server-*.noarch.rpm
on ambari-agents (3 servers used for test) :
copy ambari-agent/target/rpm/ambari-agent/RPMS/x86_64/ambari-agent-*.rpm on
each servers
with root access right : yum install ambari-agent-*.noarch.rpm
TEST :
- Deploy an empty Ambari Cluster with 3 RegionServers
- Choose Hadoop stack version >= 2.0.0
- At step "Customise Services" of installer, check Advanced parameters in
HBase tab :
* hbase.regionserver.global.memstore.size with value 0.4
* hbase.regionserver.global.memstore.size.lower.limit with value 0.95
- If you have empty settings in HBase tab clean your cache and reload page
- Change this value to 0.5 and 0.90 and finish installation
- When regionservers started check on each RegionServer
/etc/hbase/conf/hbase-site.xml :
* hbase.regionserver.global.memstore.size value must be 0.5
* hbase.regionserver.global.memstore.size.lower.limit value must be 0.90
- When installation done change parameters to 0.45 and 0.94, save changes and
restart regionservers
- check on each RegionServer /etc/hbase/conf/hbase-site.xml :
* hbase.regionserver.global.memstore.size value must be 0.45
* hbase.regionserver.global.memstore.size.lower.limit value must be 0.94
- Deploy an empty Ambari Cluster with 3 RegionServers
- Choose Hadoop stack version < 2.0.0
- At step "Customise Services" of installer, check Advanced parameters in
HBase tab :
* hbase.regionserver.global.memstore.upperLimit with value 0.4
* hbase.regionserver.global.memstore.lowerLimit with value 0.38
- If you have empty settings in HBase tab clean your cache and reload page
- Change this value to 0.5 and 0.4 and finish installation
- When regionservers started check on each RegionServer
/etc/hbase/conf/hbase-site.xml :
* hbase.regionserver.global.memstore.upperLimit value must be 0.5
* hbase.regionserver.global.memstore.lowerLimit value must be 0.4
- When installation done change parameters to 0.45 and 0.39, save changes and
restart regionservers
- check on each RegionServer /etc/hbase/conf/hbase-site.xml :
* hbase.regionserver.global.memstore.size value must be 0.45
* hbase.regionserver.global.memstore.size.lower.limit value must be 0.39
OUTPUT of mvn clean test :
- Without my patch : TRUNK-OUTPUT-TEST.txt
- With my patch AMBARI-6485-OUTPUT-TEST.txt
> hbase.regionserver.global.memstore.lowerLimit and
> hbase.regionserver.global.memstore.upperLimit are deprecated
> --------------------------------------------------------------------------------------------------------------
>
> Key: AMBARI-6485
> URL: https://issues.apache.org/jira/browse/AMBARI-6485
> Project: Ambari
> Issue Type: Bug
> Affects Versions: 1.7.0
> Reporter: Azwaw OUSADOU
> Labels: patch
> Fix For: 1.7.0
>
> Attachments: AMBARI-6485-OUTPUT-TEST.txt, AMBARI-6485.diff,
> DEPRECATED_OPTS.diff, TRUNK-OUTPUT-TEST.txt
>
>
> On new hbase version, hbase.regionserver.global.memstore.lowerLimit changed
> to hbase.regionserver.global.memstore.size.lower.limit and
> hbase.regionserver.global.memstore.upperLimit changed to
> hbase.regionserver.global.memstore.size.
> This patch changes these parameters for Hadoop stacks >= 2.0.0
--
This message was sent by Atlassian JIRA
(v6.2#6252)
