[
https://issues.apache.org/jira/browse/TINKERPOP-1091?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15107663#comment-15107663
]
Marko A. Rodriguez commented on TINKERPOP-1091:
-----------------------------------------------
See
http://lordjoesoftware.blogspot.com/2015/02/using-kryoserializer-in-spark.html
> Get KryoSerializer to work natively.
> ------------------------------------
>
> Key: TINKERPOP-1091
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1091
> Project: TinkerPop
> Issue Type: Improvement
> Components: hadoop
> Affects Versions: 3.1.0-incubating
> Reporter: Marko A. Rodriguez
> Labels: breaking
> Fix For: 3.2.0-incubating
>
>
> Right now, if you use SparkServer (cluster mode) then you need to have
> {{GryoSerializer}} on all the machines in the cluster. That is,
> {{SparkContext.addJar()}} does NOT work for {{spark.serializer}}. This is
> really lame.
> I know we want to get make {{GryoSerializer}} able to work more generally
> ([~dalaro]). Lets see if we can't get {{KryoSerializer}} with custom added
> TinkerPop3 classes to work. This way, we don't have to copy jars to machines
> and can rely on {{loadJars()}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
