<html><body>
<p><font size="2" face="sans-serif">Hello,</font><br>
<br>
<tt><font size="2">> Looking at FPO it appears to require being setup as a separate <br>
> 'shared-nothing' cluster, with additional FPO and (at least 3) <br>
> server licensing costs attached. Presumably we could then use AFM <br>
> to ingest(/copy/sync) data from a Hadoop-specific fileset on our <br>
> existing GPFS cluster to the FPO cluster, removing the requirement <br>
> for additional gateway/heads for user (data) access? At least, <br>
> based on what I've read so far this would be the way we would have <br>
> to do it but it seems convoluted and not ideal.<br>
</font></tt><br>
<tt><font size="2">GPFS FPO nodes can become part of your existing cluster. </font></tt><br>
<tt><font size="2">Have you read this document? If not, take a look, it contains quite a lot of details on how it's done.</font></tt><br>
<a href="http://public.dhe.ibm.com/common/ssi/ecm/en/dcw03051usen/DCW03051USEN.PDF"><tt><font size="2">http://public.dhe.ibm.com/common/ssi/ecm/en/dcw03051usen/DCW03051USEN.PDF</font></tt></a><br>
<tt><font size="2">Also take a look at the public GPFS FAQs which contain some recommendations related to GFPS FPO.</font></tt><br>
<br>
<tt><font size="2">> Has anyone else run Hadoop alongside, or on top of, an existing san-<br>
> based GPFS cluster (and wanted to use data stored on that cluster)? <br>
> Any tips, if you have? How does it (traditional GPFS or GPFS-FPO) <br>
> compare to HDFS, especial regards performance (I know IBM have <br>
> produced lots of pretty graphs showing how much more performant than<br>
> HDFS GPFS-FPO is for particular use cases)?<br>
<br>
Yes, there are GPFS users who run MapReduce workloads against multi-purpose GPFS clusters that contain both "classic" and FPO filesystems.</font></tt><br>
<tt><font size="2">Performance-wise, a lot depends on the workload. </font></tt><br>
<tt><font size="2">But also don't forget that by avoiding the back-and-forth copying and moving of your data isn't directly measured as better performance, although that too can make turnaround times faster.</font></tt><br>
<br>
<tt><font size="2">Regards</font></tt><br>
<tt><font size="2">Sean</font></tt></body></html>