pcp
[Top] [All Lists]

RE: [pcp] PCP and Hadoop (or any other large cluster of commoditymachine

To: "Paul Smith" <psmith@xxxxxxxxxx>, <pcp@xxxxxxxxxxx>
Subject: RE: [pcp] PCP and Hadoop (or any other large cluster of commoditymachines)
From: "Hearns, John" <john.hearns@xxxxxxxxxxx>
Date: Mon, 17 Jan 2011 09:33:22 -0000
In-reply-to: <7EE5FC98-2E3D-4429-BFD0-023CD3DA23BE@xxxxxxxxxx>
References: <7EE5FC98-2E3D-4429-BFD0-023CD3DA23BE@xxxxxxxxxx>
Thread-index: Acu19K/ZEHYaJbTMQYOqctkVRsXjggANCExQ
Thread-topic: [pcp] PCP and Hadoop (or any other large cluster of commoditymachines)
> 
> Nathan mentioned about SGI already having to deal with something like
> this with a hierarchy of pmdas, although that's about as deep a
> conversation as we got.  I could imagine a higher-level pmda sucking
> all data for an individual rack, summarising and providing it upstream
> for a cluster-level pmda.   I would really like to understand more
> about this approach.
>
Yes indeed, SGI have a heirarchical approach for using PCP on ICE
clusters.


http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi/linux/bks/SGI_Adm
in/books/ClusterMan_AG/sgi_html/ch05.html#Z1210619170tls

or http://tinyurl.com/63v9p6g

As this is part of SGI's value add when supplying and supporting
clusters, maybe they could be approached for licensing this
software to folks who run the large clusters you describe, but not
necessarily using Tempo.

The contents of this email are confidential and for the exclusive use of the 
intended recipient.  If you receive this email in error you should not copy it, 
retransmit it, use it or disclose its contents but should return it to the 
sender immediately and delete your copy.

<Prev in Thread] Current Thread [Next in Thread>