The dual hash miner seems rock solid to me, have been running for 7 days on one ubuntu rig i set up.
Other than the things we have already talked about , eg xmr charity latency not displaying i still have some bbp rejects and i suppose most of them are due to the high latency for the shares..
they average around 300-600ms the bbp shares that is.
So what do you think of setting up a european pool server Rob for when mainnet launch or too expensive?
Yeah i have no experience from pool operating or anything like that just shooting out ideas from my head.
Oh great, thats good to hear bout the ubuntu miner - as I have only tested that minimally, so Im glad its acting as reliably as the windows miner, perfect.
Regarding the xmr-charity shares, I did fix that but I need to check the code in and release it - I will try to do that today just so we can knock that out.
Yeah, it seems to me that all of the issues I can think of in the xmrig miner have been solved, and the pool has been running fine without any reboots. The core wallet is now running fine in both solo mining and (I think, the sancs are OK with chainlocks, we will check that again), but to my knowledge the only feature that is not programmed yet- that we should probably tackle before we try to launch this baby - is the pool spork enforcement reqs. As I know we will eventually need them. Ill try to put this feature in soon. This just basically allows us to enforce that blocks solved in biblepay-rx in the future must come from one of our pools - this is so people cant get around mining personal blocks to circumvent giving to charity (they will all have to pool mine if difficulty is greater than say, 10 for example in prod).
As far as multiple pools, Its going to be very easy to run, at least in this first version. Nomp can be set up on a dual core vm in about an hour. So I think it would only cost someone $20 to run one in Europe on vultr. However I would be looking for a volunteer to run other pools as Id like to limit myself to running one (but I will say this I will scale mine up as big as necessary, so if it starts to slow down due to having 250 miners Ill gladly add more nodes at that point and keep scaling it up) - I just mean, Id rather not be involved in running multiple domains myself. Btw, we give out free subdomains to anyone who wants to start a european pool - for example, europa.biblepay.org - I can point that DNS to anyone who wants it.
On the bbp rejects, I did go through the pain of trying to reduce all other types (other than stale) to the point where it appears the business logic is very accurate (another words, the rejects all point to stale shares). I believe the stales are more of a function of 'how many people are mining in testnet'. As I imagine if we had 100 miners (in contrast to 2), I imagine that 99.9% of the shares the pool gives will not be stale. In testnet we have this high propensity for receiving a stale share when: Miner has asked for 10 threads of work, someone solves the block, then for the next 30 secs, the latent threads keep submitting stale shares. In prod, we should see many more partial solutions (IE shares were submitted on low diff results - not a block solution) resulting in a higher % of non-stale, I believe.
