[ https://issues.apache.org/jira/browse/TINKERPOP-1389?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15484879#comment-15484879 ]
Marko A. Rodriguez commented on TINKERPOP-1389: ----------------------------------------------- I was able to fix the performance issue. It had to do with creating too many GryoPools due to the transient serialization as suspected. Note that {{HadoopPools}} exist for just this situation. Here are the runtimes for the process computer test suite for Spark. {code} 4:13 // GryoSerializer (prior to my last commit — 25 minutes) 4:49 // KryoSerializer {code} > Support Spark 2.0.0 > ------------------- > > Key: TINKERPOP-1389 > URL: https://issues.apache.org/jira/browse/TINKERPOP-1389 > Project: TinkerPop > Issue Type: Improvement > Components: hadoop > Affects Versions: 3.2.1 > Reporter: Chen Xin Yu > Fix For: 3.3.0 > > Attachments: TINKERPOP-1389.patch > > > Spark 2.0.0 was released: > http://spark.apache.org/news/spark-2-0-0-released.html > There are lots of improvement and changes compared to 1.6.1, we should better > bump to it for TinkerPop. -- This message was sent by Atlassian JIRA (v6.3.4#6332)