[Backlogmanager] [FIWARE-JIRA] (HELP-14416) [fiware-stackoverflow] Fiware Orion performance variations

[Veronika Vlnkova (JIRA)]

     [ https://jira.fiware.org/browse/HELP-14416?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Veronika Vlnkova resolved HELP-14416.
-------------------------------------
    Resolution: Done

> [fiware-stackoverflow] Fiware Orion performance variations
> ----------------------------------------------------------
>
>                 Key: HELP-14416
>                 URL: https://jira.fiware.org/browse/HELP-14416
>             Project: Help-Desk
>          Issue Type: Monitor
>          Components: FIWARE-TECH-HELP
>            Reporter: Backlog Manager
>            Assignee: Jose Manuel Cantera
>              Labels: fiware, fiware-orion
>
> Created question in FIWARE Q/A platform on 17-07-2018 at 11:07
> {color: red}Please, ANSWER this question AT{color} https://stackoverflow.com/questions/51377854/fiware-orion-performance-variations
> +Question:+
> Fiware Orion performance variations
> +Description:+
> I have some general questions about Orion performance, in general i have no issues with Orion and it is working properly. But i am seeing some strange patterns and need some advice about.
> So my setup is docker-compose.
> Cli commands used while creating:
> Orion -dbhost mongo -reqMutexPolicy none -dbPoolSize 100 -logLevel NONE
> Mongo --nojournal --logappend 
>   Question:
> Case:
> 3000 entities in batch operations, split into 5 + batch. So 5 + connection 
> My PC 8 gig ram, i7
> Times:
> First time creating: 
> POST 200 7341.858 ms - 5 take notice and - 3 seconds of this because my program 
> take that much to do some checks so 4 sec.
> Second time only difference service path:
> POST  200 10845.319 ms - 5
> Third time creating difference service path:
> POST  200 16033.816 ms - 5
> Next calls 20 sec, 25 sec, 33 sec, 45 sec.... Only difference between them  is service path.
> So as u can see each next request is simply taking longer ( or that is expected ), i have read documentation about performance but failed to find something similar to this.
> It seem that mongo is issue here for some reason it is inconsistent, and not by expected margin but much more.
> Picture of docker stats command:



--
This message was sent by Atlassian JIRA
(v6.4.1#64016)