On Nov 24, 2012, at 4:25 PM, Matthew Farrellee <matt@xxxxxxxxxxx> wrote:
> On 11/23/2012 01:36 PM, Todd Tannenbaum wrote:
>>
>> Just FYI, I found the following of interest:
>>
>> http://www.mesosproject.org/
>>
>> Apache Mesos is a cluster manager to provide resource isolation and
>> sharing across distributed frameworks. E.g. effectively run HTCondor,
>> PBS, Hadoop, Spark, etc, on the same cluster hardware. Uses LXC (linux
>> containers).
>
> Likewise, Corona from Facebook -
>
> https://www.facebook.com/notes/facebook-engineering/under-the-hood-scheduling-mapreduce-jobs-more-efficiently-with-corona/10151142560538920
>
> And of course YARN -
>
> http://hadoop.apache.org/docs/r0.23.0/hadoop-yarn/hadoop-yarn-site/YARN.html
>
> This space just keeps growing with clear pain points around application integration and workload consolidation.
>
As a mind game, what would it take to implement the same thing within the Condor ecosystem? What kind of interfaces could we provide to expose the matchmaking and the claim protocol to others?
Just a thought...
Brian
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
|