Hello Created a load test where we had two server nodes in our topology and
loaded with 10 gig of data.
After loading to the point where off heap max was reached we issued a clear
and emptied the cache. Verified that off heap entry counts were set back to
zero on both our servers. The heap cleared completely on one of the server
nodes the other node heap did not go down. It looks like all the memory was
in the Old Gen Space.
Here is our cache configuration
<bean id="imageCache" abstract="true"
class="org.apache.ignite.configuration.CacheConfiguration">
<property name="name" value="ImageCache"/>
<property name="cacheMode" value="REPLICATED"/>
<property name="memoryMode" value="OFFHEAP_TIERED"/>
<property name="swapEnabled" value="false"/>
<property name="indexedTypes">
<util:list id="myList" value-type="java.lang.Class">
<value>java.util.UUID</value>
<value>com.mgic.documentviewer.imaging.cache.beans.TiffPage</value>
</util:list>
</property>
<property name="sqlOnheapRowCacheSize" value="1" />
<property name="atomicityMode" value="ATOMIC" />
</bean>
<bean parent="imageCache">
<property name="offHeapMaxMemory"
value="#{10 * 1024L * 1024L * 1024L}"
/>
<property name="statisticsEnabled"
value="true"/>
</bean>
--
View this message in context:
http://apache-ignite-users.70518.x6.nabble.com/One-server-node-seems-to-hang-onto-heap-memory-after-clear-tp8838.html
Sent from the Apache Ignite Users mailing list archive at Nabble.com.
