I have a couple of servers that sit idle for periods of time (and when busy they are usually IO-busy rather than CPU-busy), so I've decided to dabble with bitcoin out of idle curiosity.

Using the standard client (v0.3.19 currently) with instances of the sample miner connected to it, it only seems to count its own work. i.e. it shows ~2,000Khps whether the miners are connecting and doing work or not. The total from all six machines, going by the output of the miner instances, should be about 12,000.

Is it normal to see just the main worker's count in the output of ./bitcoind getinfo and ./bitcoind gethashespersec, or do I have something wrong and should be seeing the larger number? If this is expected behaviour rather than due to misconfiguration on my part, is there a way to query the daemon for the total rate for itself and its slave miners?

The question may be moot of course, as with 12Mhps and difficulty set to rise noticeably I'd probably be better off using something like bitpenny if I hope to pickup any coins this year (I'm not going add in my gaming rig as I find it hard to believe that paying for the electricity to keep it running it at full tilt 24/7 and replace anything that wares out fasted that it otherwise would wouldn't be a loss making exercise - the machines I'm currently running it on are in DCs where they run 24/7 anyway and I'm not separately charged for power use).