OK, I had to remove email queue, so please resend your confirmation email again, it should be fixed now (and email will arrive when your mailserver unblacklist that IP). I'm sorry for such troubles, I'm not experienced in configuring postfix :-/.

LoupGaroux: Thanks, I found that IP of mailserver was on one blacklist. It's removed now, emails should be processed up to two hours (as mailservers refresh their blacklists).

Unfortunately one of database crash was during calculation of namecoin rewards, which made database inconsistent and I need to fix it before I enable payouts again. I'm see that some blocks were accounted to user's namecoin balance twice. Fortunately I have text logs for everything, so I need to write scripts to parse them and reconstruct rewards back to consistent state. But it is a lot of work and namecoin rewards will be paid out tomorrow, not today. I'm really sorry for that.

bulanula: NMC payouts are still stopped, but BTC payouts works. Simply change your sending threshold to 0.01 and pool will send it to you in 15 minutes.

Bitcoin payouts are done, it was around 1600 BTC. I'm going to fix nmc payouts right now.

This. And also the BTC payout won't be so huge, don't expect any market move because of this payout.

OK. Pool is up for 11 hours without any crash, which is very good sign - it's much longer than before while running on VMWare, so that issue looks fixed.

I see a little higher stale ratio on some miners. Nothing strange - around 0.8-1.3% stale, but it should be near 0.3%. It's pretty normal - I didn't expected that it will be "reinstall and forget", every environment need some specific tuning.

So everything seems stable, it's time to process payments!


Hmmmm, write me a PM, we can make a deal.

tinmann: I'll recalculate all those blocks affected by outages with proportional method. You're right that score method don't work well when miners cannot submit their shares. So I'm sure you'll have something from those blocks, too.

Last few weeks were really crazy. It started with merged mining hacking, then DDoS attack, cancelling linode, finding another provider and migration from Xen to VMWare. And finally debugging that strange disk issue and migrating from VMWare to bare metal. Of course with very long and painful outages, but sometimes many weird things happen at the same moment.

Now pool runs nicely, actually the server is idling for current hashrate. Everything I wish is to have pool stable for some time, because I simply need some rest.

Ok, I think I'm done. Anybody see any issue in mining?

I'm sorry guys, but payments must wait, I need some sleep. It's 6am here and I don't feel so fresh after 9 hours of intentsive work on pool setup.

Common guys, it's not configured yet. Be more patient, please.

...most of them don't know how to quit vim. I don't care anymore about some guys on this forum.

...and now kids, now I tell you how your grandfather went to madhouse. Few decades ago, there was one opensource project called Bitcoin...

ESX 5. It's my first (and painful) experience with VMWare (except vmware player few years ago).

And now even my main Internet connection fall down (it's second time in last 5 years). I'm working on mobile connection, making reinstall even harder. omg.

FoxMURDER: Database fails in access to disk sectors, although disk looks good. It happen only under high load so I'm expecting bug in vmware disk abstraction layer.

I'm going to remove virtualization completely and try to run pool on bare metal. However it will take some time to install and setup everything.

OK, I reinstalled database machine to different distro, different kernel, different version of database. And it is still crashing. It's 100% bug in VMWare ESX, which is pretty strange. I'm lost.

I'm here for a moment . No, this block is perfectly valid, you can click to blocknumber and see details on block explorer.

All miners over Tor? Yes. See http://torstatus.blutmagie.de for Tor routers and their bandwith... Actually *internal* speed should be pretty good (capacity of exit nodes is different story).

Pool traffic is currently around 1MB/s and it's roughly 10% of Bitcoin hashrate. So all miners over Tor would take about 10MB/s, which is still pretty acceptable.

Edit: I'm sorry, but I cannot finish processing payments, have some important meeting now. But no coins are lost, I just need some time to recalculate crippled rounds from time out pool outages.

I has been using Xen everytime for many years (on Linode as a customer, but also on my own boxes) and once it was running, I never had any issue with it. I'm also surprised that it don't work  out of the box.

OK, I misused recent pool instability to some software updates, so there were few more backend restarts not related to kernel issues. But now I'm leaving it 'as is' and will see if another kernel fixed the issue (it should, but I told it about that kernel options, too).