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 2015/04/24 22:54:38 UTC

[jira] [Updated] (TINKERPOP3-643) Websocket refuses requests after a certain number of requests

     [ https://issues.apache.org/jira/browse/TINKERPOP3-643?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

stephen mallette updated TINKERPOP3-643:
----------------------------------------
    Affects Version/s:     (was: 3.0.0.GA)
             Assignee: stephen mallette

Are you doing this against a milestone release like M7?  or did you build from master?  something else? 

And just to clarify, I understand that you created your own OpProcessor (that just returns a string of some sort)? how are you communicating with the server? 

> Websocket refuses requests after a certain number of requests
> -------------------------------------------------------------
>
>                 Key: TINKERPOP3-643
>                 URL: https://issues.apache.org/jira/browse/TINKERPOP3-643
>             Project: TinkerPop 3
>          Issue Type: Bug
>          Components: server
>         Environment: OS X, LInux 64Bit
>            Reporter: Ben Livingood
>            Assignee: stephen mallette
>
> Given a websocket connection and an op, server fails to service Websocket connections after an almost fixed number of requests. 
> Steps done so far:
> Created a simple OpProcessor to service a static string. 
> Started asking server to service said request, after ~16300 requests on a retina macbook or ~22k requests on a gce machine the websocket refuses to answer any more requests for a certain time. 
> After window for processing more requests opens up, the server will only serve a more limited number of requests (in the low thousands) before again ceasing to resolve any traffic. 
> A longer period of time waited tends to give more op's processed but is not guaranteed to deliver the same high water mark as initial run and definitely never exceeds it. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)