incubator-hama-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Thomas Jungblut <thomas.jungb...@googlemail.com>
Subject Re: [jira] [Commented] (HAMA-511) Submitting heterogenous supersteps with precedence constraints on Hama
Date Fri, 10 Feb 2012 10:01:41 GMT
Selective synchronization can be much cleaner if you treat this as a graph
of supersteps.

2012/2/10 Edward J. Yoon <edwardyoon@apache.org>

> There's a trade off I think. Simple/agile/easy-to-use vs. complex job
> optimization.
>
> If both barrier and selective sync types can coexist in a hama-core, I'm
> +1.
>
> If there's some plan to implement  acyclic graph job manager, it'd be nice
> if we create new module I think.
>
> 나의 iPhone에서 보냄
>
> 2012. 2. 10. 오후 6:05 "Thomas Jungblut (Commented) (JIRA)" <jira@apache.org>
> 작성:
>
> > We should just ask ourselfs, why they dropped dryad.
>



-- 
Thomas Jungblut
Berlin <thomas.jungblut@gmail.com>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message