Hello All, I've been running pvfs2 across 20 I/O nodes in a 60 node Linux Beowulf cluster for almost a year now and have been quite impressed with its performance. However, since there is no shared storage between the I/O Nodes, this particular file system is only being used as a scratch space. Now I'm looking at providing a HA pvfs2 file system for storing Bioinformatic data that needs a more stable environment than my current pvfs2 implementation can provide. This data currently resides on a singe NFS server which was sufficient for several years, but now the data is increasing in size and we have several new groups that want to access the data. To say this is a bottleneck is an understatement.
I am looking at purchasing a SAN for our cluster to use with PVFS2. The basic system will consist of four I/O servers connected via Fiber Channel (FC2) to a 6TB EMC Storage Enclosure. I've read through the pvfs2-ha documentation as well as other posts on the mailing list, but I was wondering if anyone else is running a similar configuration and what their experiences have been. Are there any specific hardware or software issues I should be aware of before taking the dive? Thanks for the help, --Luke -- Luke Hindman Research Assistant Computer Science Department Boise State University, Boise, ID 83725 [EMAIL PROTECTED]
_______________________________________________ Pvfs2-users mailing list [email protected] http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
