[ 
https://issues.apache.org/jira/browse/TINKERPOP-2888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17713481#comment-17713481
 ] 

Clément de Groc commented on TINKERPOP-2888:
--------------------------------------------

I wouldn't consider our traversals heavily nested and we're only using 
{{ReadOnlyStrategy}} at the moment I think (though JanusGraph has [a number of 
default 
strategies|https://github.com/JanusGraph/janusgraph/blob/v0.6.3/janusgraph-core/src/main/java/org/janusgraph/graphdb/database/StandardJanusGraph.java#L131-L142]
 applied automatically).

> DefaultTraversal's applyStrategies performance decrease
> -------------------------------------------------------
>
>                 Key: TINKERPOP-2888
>                 URL: https://issues.apache.org/jira/browse/TINKERPOP-2888
>             Project: TinkerPop
>          Issue Type: Bug
>          Components: server
>    Affects Versions: 3.6.1, 3.5.4
>            Reporter: Clément de Groc
>            Priority: Major
>
> We have recently upgraded from JanusGraph {{0.6.2}} (TinkerPop {{{}3.5.3{}}}) 
> to JanusGraph {{0.6.3}} (TinkerPop {{{}3.5.5{}}}) and have observed an 
> increase in all latency metrics reported by Gremlin Server.
> Using a profiler we have seen that more time is spent in 
> {{TraversalHelper.applyTraversalRecursively}} calling itself recursively 
> multiple times.
> Then we've narrowed down the issue to [this 
> change|https://github.com/apache/tinkerpop/commit/1a548e808922a0eae23b586fe4d6567238989299]
>  and resolved the latency problem by reverting the change in 
> {{{}DefaultTraversal.java{}}}.
> TLDR; as far as I understand, it looks like, in our case at least, this 
> [other round of recursion does have a significant performance 
> impact|https://github.com/apache/tinkerpop/pull/1699#issuecomment-1170430130].
>  As I'm not too familiar with that part of TinkerPop code, it's unclear to me 
> if that's because of our usage of JanusGraph/TinkerPop (feel free to advise). 
> But in any case, I thought that was worth reporting in case others hit the 
> same issue.
> Happy to share/investigate more if needed.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to