You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tinkerpop.apache.org by "stephen mallette (JIRA)" <ji...@apache.org> on 2017/05/23 19:58:04 UTC

[jira] [Commented] (TINKERPOP-1676) Improve GraphSON Performance

    [ https://issues.apache.org/jira/browse/TINKERPOP-1676?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16021736#comment-16021736 ] 

stephen mallette commented on TINKERPOP-1676:
---------------------------------------------

Already pushed this little change to make better use of memory:

https://github.com/apache/tinkerpop/commit/7ea38751871f6eddd4c4bcfe8ac82300a08f5a7d

It basically reads vertices/edges from a stream in TinkerGraph serialization (when passing subgraphs back from Gremlin Server) rather than building up a big List of vertices/edges in memory to later iterate through. That change seemed to reduce the number and frequency of GC operations.

> Improve GraphSON Performance
> ----------------------------
>
>                 Key: TINKERPOP-1676
>                 URL: https://issues.apache.org/jira/browse/TINKERPOP-1676
>             Project: TinkerPop
>          Issue Type: Improvement
>          Components: io
>    Affects Versions: 3.2.4
>            Reporter: stephen mallette
>            Assignee: stephen mallette
>
> Not sure what will happen here. Just wanted an issue where I could experiment a bit with and gather feedback from others for collaboration in case others had ideas. Right now, GraphSON reads are considerably slower than Gryo and writes appear to be faster, so the focus will be on how to get reads more in line with what Gryo can do. 



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)