Does availability trump capacity when calculating ratings?
Say I have 2 computers, would I be better off earning points by running one VM (running the Maidsafe software), and having the 2 nodes set up with DRBD so they never go down as machines reboot?
Or would I earn a better rating by running 2 VMs, and just shutting one down for a little if a machine needs a reboot?
Ideally, I’d have 3 nodes, run 2 VMs, and live migrate whichever VM to the spare node when a machine needs a reboot. But there’s going to be a gap between the time I have 2 machines & 3. Assuming availability is factored with a high weight.
I’ve been thinking a lot about these things last week, and I’ve simply termed that the best action is to just wait until the testnet where farming will be tested to test your rigs/setups. Also, I feel like it will depend on usage of the network, if there is not much network activity and the rig is not at full service then also the reward will be smaller.
At this point, we are unable to provide any figures about mining speed, there are simply too many variables to even attempt to calculate this with any accuracy. However, as we get the test networks up and running through the course of the next few months, we will start to publish this test data which may lead to a clearer picture about what users can expect. As the network stores data at random in a lottery system, it is likely that if you are looking to farm on a more commercial scale you will be more efficient using multiple commodity PCs (or NAS devices) rather than a single, high powered farming rig.
So even if you go and load up on laptops, you might be less efficient cost wise than virtualization, or a well thought out tower build.