I think you'd be hard pressed to break a couple of gb a month with a 300mh/s miner.
Awhile back I was running ~4.5gh over a 3g connection (and capped it out after about 3 weeks). I've since gotten a real broadband solution that doesn't care about usage. But hey my farm is much larger now.
Topic: [Q] pool mining data transfer (Read 749 times)
Thanks for the reply. It's helpful. I was/am actually interested in several things:
- I am supposed to get one of those USB miners pretty soon. It's much sooner than I had anticipated obtaining any mining gear. So, I'm delinquent in developing a good understanding of the different software and options. I have a cable modem in one home and a satellite connection in another, the latter being capped at 15G/mo, so this is a factor in what I decide to do where. I have actually not decide whether to even try to mine with it. It's as much a memento (and possible vanity address generator) as anything for me.
- More generally I would like to try to project how the mass availability of convenient 300MH/s-ish devices might play out in different scenarios, and how it might impact the participant's feelings about Bitcoin growth strategies.
- I am supposed to get one of those USB miners pretty soon. It's much sooner than I had anticipated obtaining any mining gear. So, I'm delinquent in developing a good understanding of the different software and options. I have a cable modem in one home and a satellite connection in another, the latter being capped at 15G/mo, so this is a factor in what I decide to do where. I have actually not decide whether to even try to mine with it. It's as much a memento (and possible vanity address generator) as anything for me.
- More generally I would like to try to project how the mass availability of convenient 300MH/s-ish devices might play out in different scenarios, and how it might impact the participant's feelings about Bitcoin growth strategies.
Probably the wrong forum... but I'll take a shot at it.
Solo mining - you run your own node and bitcoind handles block validation and submission. Bandwidth is whatever it is for a node running.
Pooled mining is the same thing except instead of getting work from bitcoind (server or on same machine) you're getting it from the pool server. You don't need to run a node yourself.
Stratum is apparently much nice to bandwidth than getwork or getblocktemplate. I think it does something like what you've described (passing a generated merkle branch or root to the miner).
What exactly is your goal? Simply to understand how it functions? If you're trying to solve a specific problem like data transfer limits, connect speed issues or something else... being more specific will get you better answers.
Solo mining - you run your own node and bitcoind handles block validation and submission. Bandwidth is whatever it is for a node running.
Pooled mining is the same thing except instead of getting work from bitcoind (server or on same machine) you're getting it from the pool server. You don't need to run a node yourself.
Stratum is apparently much nice to bandwidth than getwork or getblocktemplate. I think it does something like what you've described (passing a generated merkle branch or root to the miner).
What exactly is your goal? Simply to understand how it functions? If you're trying to solve a specific problem like data transfer limits, connect speed issues or something else... being more specific will get you better answers.
I've looked around at some documentation, but could not really spot any answer to my question. And I've never mined. I'm wondering how much data, compared to the normal peer2peer transaction data, a miner needs to solve hashes when working for a pool?
If there is a reduction in data size, it is a matter of
- the pool operator handling and discriminating out unspent or invalid transactions or other cruft?
- the pool operator handling some sort of optimization of datastructures while all the transactions are actually seen by the miner in some form?
- the pool operator computing Merkle hashes for sub-sets of the data and handing them over or something like that?
Edit: Shit! wrong board for this post. Mining support or software would have been better. Oh well.
If there is a reduction in data size, it is a matter of
- the pool operator handling and discriminating out unspent or invalid transactions or other cruft?
- the pool operator handling some sort of optimization of datastructures while all the transactions are actually seen by the miner in some form?
- the pool operator computing Merkle hashes for sub-sets of the data and handing them over or something like that?
Edit: Shit! wrong board for this post. Mining support or software would have been better. Oh well.
Jump to: