Sorry about that.  Back when BTC Guild started getting DDoS'd about 2 months ago, I gave many users a private server they could access to mine from during the attacks.  The server was hosted outside of the main datacenter, and as a result was not hit when the other servers were.  Many users were moved to another server after the first round of attacks, still private and "protected" (due to anonymity), but not at the same location.  ASICMINER was still mining on the original protected/private server.

This week I have moved all servers back into either colocation or Amazon EC2, to get away from servers with such open security systems (no 2-factor auth, poor/no IP whitelisting, too much control available through the management panels, etc.).  As a result, ASICMINER had to shift their Stratum proxies to a new IP, and it caused a temporary dip in the hash rate.  However, the result of the move is ASICMINER is now hashing on a colocated server in a very stable and reliable datacenter, one that will likely never be moved as long as BTC Guild is in business.

Merged Mining should be stable again shortly.  I was up til 3 AM trying to get it to handle the new load, and at the time I went to sleep it seemed to be doing so.  However, sometime during the night it failed and completely stopped handing out work.  It's back up now and I believe I've been able to identify what caused the major problems last night with restarting the server with so many users trying to connect at once.

At this point, I would continue to encourage miners that are stuck on getwork to consider other pool options.  At this time I simply cannot afford to toss extra servers at the getwork protocol to handle the number of users trying to utilize it.  There are plenty of other pools out there that are willing to dedicate more time to maintaining getwork based poolservers.

Sorry for the long reload of the getwork servers.  The restart process of the server is very slow due to worker authorization + prioritization of already authorized workers.  It means after the first few get in, it gets slower and slower at letting new workers get authorized.  As a result, restarting the server is taking a significant amount of time for some users to resume working.

I've made a few tweaks to the configuration, but I'm limited on how much I can do with getwork.  This is one of the reasons I encourage all users to upgrade to Stratum.  With Stratum, I designed the pool software myself, I can find and easily fix bottleneck issues.  With getwork, that is not an option, I can only fiddle with a few settings and make tweaks.  Major overhauls of the design are not possible since I am extremely limited in my knowledge of Java.  Miners get the side benefit of Stratum being a significantly more efficient protocol as well.

I have no plans to rewrite my Stratum code to add dependency on namecoind (which is extremely outdated) and rewrite the coinbase creation to add merged mining.

Getwork (regular and merged mining) servers are being restarted to try to force miners to connect to the proper servers after some migrations and DNS changes.  It should be available again within the next 30 minutes, hopefully faster!

Upcoming Changes To Payouts

As difficulty rises and still more miners come online, it is becoming very frequent for members to request payouts that are extremely small in size (under 0.10).  As a result, the pool is constantly churning through coins which have had confirmations, reducing the priority of the transactions, which means they often take longer to receive the first confirmation.  When the network has bad luck, it is sometimes the case where the pool has 30 or 40 payouts without confirmations, and some of them are waiting for a handful of other transactions which are the source for their inputs.


Later this weekend, small payouts (less than 0.10) will no longer be available for Instant Withdrawal.  Instead, these payouts will enter 'Pending' status.  Every 30 minutes, pending payouts will be grouped together and sent as a single "sendtomany" payment, similar to what is done with automatic payouts already.

Please note that for very small payouts, it is not rare for you to have to wait 30-60 minutes for the first confirmation in the first place, so this change should not have any significant change to confirmation times of payouts, and may actually improve the speed of confirmations for all users (both small and large) due to not eating through as many pre-confirmed coins.

Please provide details on what is wrong when posting something like this.  I can't really do anything when I have no idea what you're referring to.

Are you mining on FPGAs?  I've found that many FPGAs submit a significant number of dupes.  However, they are simply resubmitting a share more than once, rather than wasting time on duplicated work, and their accepted share rate still matches the expected speed.  Different miner versions have also had more or less (cgminer/bfgminer on ztex chips seem to bounce between no dupes and lots of dupes with each version update).

I apologize for EU Stratum having issues in the last 24 hours.  I've migrated eu-stratum.btcguild.com to another server in the last 15 minutes, and hopefully the new server will be stable while the DNS is pointed there.  It is temporarily not in the EU unfortunately.  However, I will be moving it back overseas as soon as possible.

I've fixed a small problem with automatic payouts.  After enabling lower amounts to be included in automatic payouts, the code was hitting 50 payouts at once, which was a limit I had hardcoded into the automatic payouts.  I've split how the automatic payouts are processed to give priority to larger payouts, but also set it to run more than once per hour if it hits a backlog.

As a result of this fix, some users may have noticed they received automatic payouts at an odd time, or multiple payouts for users who hadn't had a payout processed automatically in a while.  Going forward things should be functioning normally .

As posted above, it's hopefully coming back up soon.  Resync of the blockchain is almost caught up.


UPDATE: EU Stratum is back online.

Just adding:  BTC Guild has STILL not reached 40% of the network blocks over the last 2016.  It has been over two weeks since the first posts about BTC Guild being at/over 51%, and it would still require the pool to have 25% more hash rate than it currently has before it is true.

EU Stratum is going down, will be back up as soon as possible.


UPDATE:  EU Stratum should be restored within 1-2 hours.  A second EU Stratum server will hopefully be coming online early May.

Pools using Stratum and GBT have a fully exposed coinbase transaction which shows miners where the block will be paid.  Watching that address would reveal all blocks the pool finds.

EU Stratum server is being rebooted, should be back up in a few minutes.  Sorry for the lack of notice.

Latency still matters since stales are caused by the latency between a miner and the pool (and the pool software to some degree).  Additionally, a pool that is more globally distributed is able to push their blocks across the network faster, since having interlinked pools in multiple countries/continents guarantees the pool's block will get sent to different parts of the world faster than relying on the peer-to-peer relay, reducing the frequency of orphaned blocks.

Pools that offer true PPS are forced to run with larger hot wallets, unless you expect a pool operator to sit there refreshing a page 24 hours a day to constantly be at the ready to refill the wallet.  Proportional/PPLNS/DGM/etc pools can run with smaller hot wallets because payouts are capped at what the pool makes in the period.


I'm very sorry to hear what has happened to Ozcoin in the last few days.  For those moving to other pools, please give consideration to pools like Slush, BitMinter, or Eligius.  Those pools have been around as long or longer than Ozcoin, and have built up a strong history.  There are also many other pools available if you don't want to join a large pool.  Please do not add to the network imbalance by adding more speed to BTC Guild.

Just to point it out, BTC Guild is still in the 36-37% range out of the last 2016 blocks.  Luck has been terrible the last week and a half, so it's probably a bit understated right now, but I continue to trust http://blockorigin.pfoe.be and use that for the 51% mitigation steps that kick in at 40%+.

Website back up and out of maintenance mode.  Some lag is expected right now as a few cleanup scripts are being run and major "on the hour" scripts are running at the same time.

Somewhat related issue to the DB maintenance on the charts history:  The PPLNS share collection script got stuck and did not close shift 5062 at the appropriate interval.  The shift is being closed right now.  This shift will have roughly 2.5x as many shares as a normal shift, and will receive proportionally more of the block rewards over the shift maturation time as a result.  Sorry for the error, I've just been going through a lot of scripts again trying to keep the site moving quickly even when the heavy duty scripts are running in the background.


EDIT:  Website temporarily offline while this script is cleaned up.