Hi,
> I am not getting the exact difference between Fail-over
> clustering and Load-balancing clustering. Some web sites they are
Fail-over: When one node in your cluster fails, the other takes
over. Usually there is a heartbeat program which monitors the other
node periodically and the failover happens when something abnormal
happens (like say, the node not responding).
Load balancing: You have too much computing to do and would like to
share it between the nodes on your cluster. Example: A round-robin
DNS server. Each times it receives a request, it gives the next IP
in the machine pool so that the other machine handles the request.
Though this is not of much use in case of a caching DNS setup, hope
you get the idea of load balancing.
> giving both are same. I want to know the exact difference between the
> load balancing,failover & high performance clustering with supported
High performance: Just what it says. Lot of computing power by mixing
many machines and using it for computing. Usually for resource intensive
operations like those of computational fluid dyanmics, simulations etc.
where you havin many parallel operations would benifit. PVM libraries
etc. etc. are made use.
One or more of these could be combined together in a cluster.
It would've been good if you searched over the net for these defintions.
They don't have much to do with Linux per-se, so technically speaking,
these are way too OT here. A question on Beuwolf ... yes! a right topic
for discussion on these lists.
> softwares.
Search over the net please.
Giridhar
--
Y Giridhar Appaji Nag Giridhar.Nag[(a)t]ubinetics(.)co.in
Software evolves. It isn't designed. The only question is how
strictly you _control_ the evolution, and how open you are to
external sources of mutations. And too much control of the
evolution will kill you. Inevitably, and without fail. Always.
In biology, and in software. - Linus
an individual are not necessarily those of UbiNetics