[
https://issues.apache.org/jira/browse/HAMA-460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13173004#comment-13173004
]
Edward J. Yoon commented on HAMA-460:
-------------------------------------
Just my note:
Why BSP?
Today, many practical large processing applications requires a good
scalability, a high compatibility with existing data systems e.g., HDFS, HBase,
and a communication capability which allows to exchange information using
message-passing paradigm beyond MapReduce.
Here, Bulk Synchronous Parallel (BSP) model fills the bill nicely and besides,
it has several main advantages over MPI:
- Provides a simple and small API
- Enables to perform better than MPI for communication-intensive applications
- Guarantees impossibility of deadlocks or collisions in the communication
mechanisms, and
- The possibility to evaluate the computational cost of an algorithm
> Website update for 0.4 release
> ------------------------------
>
> Key: HAMA-460
> URL: https://issues.apache.org/jira/browse/HAMA-460
> Project: Hama
> Issue Type: Improvement
> Components: documentation
> Affects Versions: 0.3.0
> Reporter: Edward J. Yoon
> Assignee: Edward J. Yoon
> Fix For: 0.4.0
>
> Attachments: web01.patch, web02.patch
>
>
> if you have some items to be added, please comment here.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
