A cluster of V120s - Newcomer basic questions
Hi Folks,
I've been reading the Sun Cluster manuals for the last week and a half, but still haven't got a firm grip on some of the concepts, so I'd like to bounce some ideas of you more experienced folks to see if what I'm trying to do is even workable.
I'm trying to upgrade a small-ish system from its current ramshackle basis of disparate hardware to something more concise and organized, but as is often the case, I have a small budget. What I do have, though, is about fifteen V120s sitting in the warehouse. What I'm looking to do is build a cluster of ten V120s, which would be the processing core of the network. All sunray, db, filestorage and email gateway functions would be handled by other systems, this cluster would simply run the hundreds of little scripts we use to generate reports, deliverables and the likes, no major software applications needed.
On the surface it seems like a simple idea, the V120s have dual 10/100 ethernet ports, one for the private network, one for the public. The cluster would NFS mount files from my two file servers, scripts would access the db, output files to email or the file server etc.etc., but the Sun documentation keeps talking about needing dual backside network interfaces, lots of redundant paths to disk storage etc. and I'm getting a bit lost. Is it possible to create a simple cluster of redundant systems without all of the dual path requirements?
Said another way, is this idea of a very simple stack of V120s practicle? I would really like it if my users can log onto the Sunray servers, ssh over to the cluster, write/edit their scripts, toss them into cron and let the cluster munch on them.
Would appreciate any insight and guidance you folks can provide.
Thanks
Simon
