Turns out someone has done the hard work already: testnet-in-a-box https://github.com/freewil/bitcoin-testnet-box/tree/master/1/testnet3 - so I'm just posting this for anyone else who may be in a similar situation and doesn't want to trawl Google.

I have recently setup a small private farm of three Bitcoin 0.83 daemons on an internal testnet, all connected to each other only and all set to generate Bitcoins. The problem is after 5 hours and ~100% CPU usage, not one of the instances has managed to generate a single block:


bitcoind -datadir=2 getinfo
{
    "version" : 80300,
    "protocolversion" : 70001,
    "walletversion" : 60000,
    "balance" : 0.00000000,
    "blocks" : 0,
    "timeoffset" : 0,
    "connections" : 1,
    "proxy" : "",
    "difficulty" : 1.00000000,
    "testnet" : true,
    "keypoololdest" : 1377192724,
    "keypoolsize" : 6,
    "paytxfee" : 0.00000000,
    "errors" : ""
}

They've never seen the real test net and their databases were freshly initialized, but I have noticed the difficulty seems slightly at odds to some historical posts (on here) which showed a difficulty of 0.125 for default earlier versions (testnet2 I believe). Any pointers as to exactly why exactly this might be the case and if there is any way to reduce the difficulty so some blocks start being created?

I wasn't aware Blockchain.info provided a merchant service...? Wouldn't you be better off opting for an existing payment provider E.g. Bitpay or WalletBit that provides the necessary code snippets and button graphics out-of-the-box?