Home
| Calendar
| Mail Lists
| List Archives
| Desktop SIG
| Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Linux Links | Bling | About BLU |
This is definitely compute cluster stuff, not HA stuff. The only place where you'd even want a HA system is in the cluster manager. My go-to for compute clusters these days is HTCondor. It's extremely flexible both in scalability and what you can actually do with it. It's easy to create a pool and almost as easy to flock it with other pools. And it can run just about anything including MPI applications with little or no changes. You can do pretty much anything for HA on the cluster manager. If you have a Xen infrastructure then you could do Xen live migration with DRBD to handle failover. I have something along this line planned for future enhancement of my HTCondor pool. -- Rich P.
BLU is a member of BostonUserGroups | |
We also thank MIT for the use of their facilities. |