i've been facing a lot of I/O bottlenecks since I am using a single Linux fileserver. I've recently setup a hadoop cluster (27 nodes) to re-mediate the problem. I currently have a wrapper to copy the file over to my startd node and do my crunching but I was wondering if anyone has gotten hdfs and condor to play nicely by making it seamless (without a wrapper).
--
---
Get your facts first, then you can distort them as you please.--
![[Computer Systems Lab]](http://www.cs.wisc.edu/pics/csl_logo.gif){kind=logo}