Alright, I've been looking over the documentation that was added to the Bitcoin wiki recently regarding implementation of merged mining.  At this time, the only planned alternate currency to be added is NMC.  Others could be added later, but I will be taking this slowly.

NMC will not have as much presence in the UI.  BTC Guild is a Bitcoin mining pool, and throwing more graphs/tables of data for alternate currencies is not something I plan to do.  The current plan to keep NMC implementation simple and clean is to bootstrap it to the PPLNS system.  When a namecoin block is found, it will get distributed evenly to the open PPLNS shifts the same way BTC is distributed.  This means it will not be available via PPS.  Still working on how I will handle PPS for NMC going forward.

Once the stratum code is tested and working, the getwork based merged mining server will be closed down after a reasonable period of time for users to move to Stratum.

No guarantees on a time table, but I'm setting time aside tomorrow to begin working on a branch of the BTC Guild stratum pool server that will implement merged mining.  Technical details regarding how to implement merged mining have finally been added to the Bitcoin wiki.  Until that was added, the only guidance I've found were reverse engineering the mergedmining proxy that some pools used in the past, or trying to look through the namecoin source and working backwards from there.

I dunno, a fried cat sounds like at least one life has already been lost.

Gah, for the record, the pool is NOT at 40 TH/s.  Small bug in some of the worker graphing/speed calculations from a worker I manually restored for testing.

Nothing against it, though I'm not sure why you'd need separate accounts to do it.  One word of warning:  Automatic Payouts don't work properly if you have two accounts sharing the same wallet address and triggering a payout at the same time, it requires a manual correction to fix.

Check Your Miners

The servers had a brief restart earlier today to add some changes to my botnet detection since I had a few false positives get flagged (fringe cases where legit use looked like a botnet).  The checks didn't have a way to whitelist accounts, so I had to add that into the software.  The downtime was less than 10 seconds in total.

The pool did lose a chunk of speed (relatively small, but more than 100 GH/s) that didn't come back after the servers were restarted, so if you're a BTC Guild user I would advise checking your miners just to make sure they didn't get lost during the restart.  If you had Idle Warnings enabled on the settings, an email would have been fired off if this was the case.




Yes.  I'm slowly working on a better way to visualize it and incorporate it into the page.  The last thing I want is another icon on the bar right now.

To add to this:  It took months for difficulty to drop after the 2011 crash.  Difficulty actually kept on climbing for a while.  There were significant chunks of time where most miners were mining at a loss.  Many miners do not re-evaluate their profitability as often as you think, and they frequently forget indirect costs [air conditioning is a hidden cost that really screws up most miner projections when looking at larger farms].

You can expect that the rate GPU users turn off will be so staggered out due to bad math/persistence/hopes&dreams that it won't even be noticed due to the ASIC power coming online simultaneously.

It's still a work in progress.  It's tough to get proper stats to generate automatically because shifts can include two different difficulties.  I am definitely going to add back a Last X shift average once I make this look nicer.  Currently I've manually put in which shift to start with so this only includes shifts which included only blocks at the current difficulty.

The joy of selection bias.  PPLNS paying above PPS for the first couple days after a change, no mention.  FIRST day that the last 100 shift figure drops under PPS and people complain.

Under PPLNS, not possible [PPLNS doesn't record shares per shift by worker, only by user].  Under PPS it's *possible*, but not in any way using the data on the website, and I can't imagine a way I can squeeze per worker PPS earnings in there.

Not showing any problems.  All servers are still running with a couple thousand connections each, no errors in any logs, and no drop in the pool speed to indicate a regional issue.

Not unexpected.  We had a few very good days.  We're quite ahead on luck since the difficulty change [finally] .



Actually, my alarm is humorous.  It's the Steve Balmer "Developers, Developers, Developers" rant, with bad techno in the background and includes the screaming.

2 minutes to wake up -and- make sure it wasn't a false alarm!


A small update to the situation:  Your dashboard speeds will be artificially low due to the gap in shares.  I'm updating the dashboard temporarily to try to ignore that gap so users can see that they're hashing normally again.  As a result, DISPLAYED speeds might fluctuate a little wildly for a few minutes!  Your actual speeds are obviously not affected .

Was just woken up by my alarms, looks like something is hitting the servers.  Working on figuring out what's going on ASAP.


UPDATE:  For the first time in quite a while, this was not an attack.  One of tables on the database crashed and the servers were unable to post to it.  I've repaired the table and restarted the DB server.  On the bright side, this was causing the Stratum servers to crash, allowing users to fall back to their backup servers if their miners were configured to do so.



Based on my miner output, it looks like the total outage time was around 10 minutes [5 minutes to trigger my alarm, 2 minutes for me to wake up and check it wasn't false, 3 minutes to fix].

Actually, I missed the part about altcoins.  As far as I'm aware, you can't swap cgminer from SHA256 to SCrypt altcoins without restarting it, so my solution wouldn't work anyways.

If it wasn't for Windows, I'm sure a quick shell/python script could be written to ping a central website for a signal to restart cgminer with different arguments.  It's possible with Windows too, but would be much easier in a Linux system.

If the pools all use Stratum, you could run a Stratum proxy on a central server.  Then whenver you want to move, just restart the proxy with a new pool address.  The miners should reconnect and now start sending work at the new pool.  You'll get a few rejects due to the miners submitting work from the prior server, but it shouldn't last very long.

1) It should not affect hash rate directly, but being on a different server (different location) it may have a minor affect either + or -.
2) No, merged mining on BTC Guild is only available on getwork at this time.
3) Yes, the mergedmining.btcguild.com:8332 getwork server records BTC as PPLNS while recording NMC as 5% PPS.
4) Being on getwork, it requires more bandwidth.  It is also not software I designed, so support and updates are very limited.
5) No additional fees other than the fee taken from NMC (5% PPS).

Hmm...surprisingly I -do- have it saving the worker ID of which worker found the block [thought I only kept User ID].  I'll see if I can squeeze it in somewhere (no promises).

Yep, the dry spell ended and we got a burst of 5 blocks fairly fast!


Notice to users on getwork using the default btcguild.com:8332 - The server running that pool is going to be down for maintenance tonight.  I have moved DNS over to a backup server, and you should start being redirected.  Sorry for any downtime caused.  Depending on your miner behavior and DNS settings you may also see a number of 'Unknown' rejects show up since your miner may have work stored from the old server and then submitting it to the new server.

DNS will be returned to the normal getwork server after approximately 12 hours.

For PPLNS, balances updates every single time a block has been found.  This is *normally* around every 30 minutes (sometimes much faster, sometimes slower).  While we've had amazing luck for the last day and a half, we're currently on a dry spell with ~2.5 hours [as of this post] since the last block, meaning no balances have updated in 2 and a half hours.