NFS file server
i require a load-balanced (to prevent single point-of-failure, and for high-availability) storage cluster with multiple nodes each with around 600gb of usable storage.
i want to be able to then have a few nodes, say 6, and be able to combine the storage to build an uber storage array, i also want to however have it load-balanced for the reasons listed above but i want it so that if a node does fail the data on that node is availbable else where.
how can i go about this? what are the best practices for this?
to summarise: I want the NFS cluster to have multi-terabyte file storage that is fail-safe (so that if one nodes fail the data can still be accessed) and load-balanced (so that under high load it can cope).
how does google and similar companies do this (i.e. how does gmail store our email so that it is constantly available, backed-up and multi-terabyte storage)?
thanks yet again,
josh
