Thanks Dmitry,
Will <property name="onheapCacheEnabled" value="false"/> meaning the data
will be stored off heap; so the GC should no affect it ? also the off heap
size control by the memory policy right?
<property name="memoryPolicyName" value="Trans_Region"/>
<bean class="org.apache.ignite.configuration.MemoryPolicyConfiguration">
<property name="name" value="Trans_Region"/>
<property name="initialSize" value="#{8L *1024L * 1024L * 1024L}"/>
<property name="maxSize" value="#{12L *1024L * 1024L * 1024L}"/>
<property name="pageEvictionMode" value="RANDOM_2_LRU"/>
<property name="swapFilePath" value="/var/tmp/qa/swap2"/>
</bean>This confuse us a lot, as eventually our database size is less than 1G,
so either the onheap or offheap should not exceed the limitation . The GC
should be health even the full GC is done at 100ms, frequent is about once per
half hour. Our use case is there are about 3k TPS, which may update the Cache
more than 3 times each request, so about 10k transaction happen, those data be
written bahind to the DB by batch size as 4096;So I not sure this is the
reason? BTW we use the cache totally as LOCAL. so in theory suppose no
network transfer should be involved.
Thanks again for your time!
Regards
Aaron
[email protected]
From: dkarachentsev
Date: 2017-08-15 14:11
To: user
Subject: Re: In which scenario the ignite nodes will shutdown by itself
Hi,
Node may stop if it was segmented out of the cluster. Usually this happens
because of long GC or network pauses, when node doesn't respond longer than
failureDetectionTimeout.
Please, attach Ignite (with -DIGNITE_QUIET=false JVM flag) and GC logs, this
will help to understand more.
Thanks!
-Dmitry.
--
View this message in context:
http://apache-ignite-users.70518.x6.nabble.com/In-which-scenario-the-ignite-nodes-will-shutdown-by-itself-tp16192p16193.html
Sent from the Apache Ignite Users mailing list archive at Nabble.com.
