Hi,
I have been present a problem in the forum,the link is:
http://apache-ignite-users.70518.x6.nabble.com/about-mr-accelerator-question-tc3502.html
I will follow you suggestions on the ignite cluster, but i am so sorry that
there are some confusions.
on the single node,with 29G data, my ignite configrution is default almost.just
modify the ignite.sh
as follow:
if [ -z "$JVM_OPTS" ] ; then
JVM_OPTS="-Xms8g -Xmx8g -server -XX:+AggressiveOpts -XX:MaxPermSize=8g"
fi
and modify the default-config.xml
<property name="fileSystemConfiguration">
<list>
<bean
class="org.apache.ignite.configuration.FileSystemConfiguration"
parent="igfsCfgBase">
<property name="name" value="igfs"/>
<property name="metaCacheName" value="igfs-meta"/>
<property name="dataCacheName" value="igfs-data"/>
<property name="ipcEndpointConfiguration">
<bean
class="org.apache.ignite.igfs.IgfsIpcEndpointConfiguration">
<property name="type" value="TCP" />
<property name="host" value="127.0.0.1" />
<property name="port" value="10500" />
</bean>
</property>
<property name="secondaryFileSystem">
<bean
class="org.apache.ignite.hadoop.fs.IgniteHadoopIgfsSecondaryFileSystem">
<constructor-arg name="uri"
value="hdfs://localhost:9000"/>
<constructor-arg
name="cfgPath"><null/></constructor-arg>
<constructor-arg name="userName"
value="client-user-name"/>
</bean>
</property>
</bean>
</list>
</property>
the result is very cool.
the on-heap just 8G but the data have 29G.It work well.why?
another problem is I start up another ignite node on the same server.I have 16g
heap with two ignite node.My job is finished after 7 minutes, the effect is the
same as on the single node.why the performance is not improve when add the
memory?
thanks