-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/8135/#review13617
-----------------------------------------------------------



src/webui/master/static/controllers.js
<https://reviews.apache.org/r/8135/#comment29191>

    I'm glad to see that your times are all in the same order of magnitude as 
mine, but I'm dropping this since I think your times are a bit on the low side.
    The webui locks up during update for clusters O(1000) and getting that 
every 10 seconds is very obtrusive.
    
    In the long term, we will be refactoring the webui to scale better (in 
that, we will not be polling a gigantic state.json endpoint frequently). This 
way, we can have fast updates for for statistical information (like resource 
usage), and slow updates for big endpoints (like state.json).
    
    Also, I will be revisiting these values when I implement paginated tables 
in the webui. Right now it's also slow due to the huge amount of DOM 
manipulation on each update.
    
    I hope this clarifies things a bit.


- Ben Mahler


On Nov. 20, 2012, 3:38 a.m., Ben Mahler wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/8135/
> -----------------------------------------------------------
> 
> (Updated Nov. 20, 2012, 3:38 a.m.)
> 
> 
> Review request for mesos, Benjamin Hindman and Vinod Kone.
> 
> 
> Description
> -------
> 
> The new webui is really slow on large clusters, given we poll the master 
> every 3 seconds and update all the front end data.
> 
> This is a simple fix to use an update interval based on the size of the 
> cluster.
> 
> 
> Diffs
> -----
> 
>   src/webui/master/static/controllers.js 
> dcb947175898c1d94837705abe21fab90d2addf3 
> 
> Diff: https://reviews.apache.org/r/8135/diff/
> 
> 
> Testing
> -------
> 
> local webui runs
> 
> 
> Thanks,
> 
> Ben Mahler
> 
>

Reply via email to