Topic: [CLOSED] BTC Guild - Pays TxFees+NMC, Stratum, VarDiff, Private Servers - page 327. (Read 903150 times)

no need to prove it or get charts out, I was just worried I would have to switch pools. I want to stay with BTCG, and thanks!

Running at a higher difficulty (like 4) would not have a significant affect on GPU users.  You'd have more variance, but over the course of a week you would hardly notice.  I'd need to find organofcorti's charts to get a better idea of exactly how much variance it would introduce.  However, it is not a LOSS of income, it is simply increasing the range you will see on any given day.  You're just as likely to see more income as you are to see less income from the variance.

I'm not making any changes yet, and there will be enough warning before it does happen.  BTC Guild has one of the most aggressive vardiff implementations around I believe, and it is extremely quick at moving ASICs into appropriate difficulty levels, which reduces the need to update pool-wide difficulty for non-ASIC miners.

@eleuthria:
getting back this difficulty thing:
"....3) Right now miners can set a minimum difficulty.  What I was referring to is eventually moving the entire pool to a higher minimum difficulty (probably 4 to start).  This probably won't happen until I can remove getwork from the pool."

Am I reading this right? will my GPU miners be able to accept shares at this difficulty? Is this going to force me to find a different pool? I have ASICs on order but not expected until mid-late summer.

probably something here about the law of unintended consequences..  

was able to edit poclbm.ini file to get back mining.

2013-03-21 10:31:13: Running command: poclbm.exe zif33rs_1:[email protected]:3333 --device=0 --platform=0 --verbose -r1 -v -w128 -f60
2013-03-21 10:31:13: Listener for "1" started
2013-03-21 10:31:15: Listener for "1": stratum.btcguild.com:3333 21/03/2013 10:31:15, checking for stratum...
2013-03-21 10:31:15: Listener for "1": stratum.btcguild.com:3333 21/03/2013 10:31:15, started OpenCL miner on platform 0, device 0 (Tahiti)
2013-03-21 10:31:15: Listener for "1": stratum.btcguild.com:3333 21/03/2013 10:31:15, no response to getwork, using as stratum
2013-03-21 10:31:15: Listener for "1": stratum.btcguild.com:3333 21/03/2013 10:31:15, Setting new difficulty: 1
2013-03-21 10:32:08: Running command: poclbm.exe zif33rs_2:[email protected]:3333 --device=2 --platform=0 --verbose -r1 -v -w128 -f60
2013-03-21 10:32:08: Listener for "2" started
2013-03-21 10:32:10: Listener for "2": stratum.btcguild.com:3333 21/03/2013 10:32:10, checking for stratum...
2013-03-21 10:32:10: Listener for "2": stratum.btcguild.com:3333 21/03/2013 10:32:10, started OpenCL miner on platform 0, device 2 (Devastator)
2013-03-21 10:32:10: Listener for "2": stratum.btcguild.com:3333 21/03/2013 10:32:10, no response to getwork, using as stratum
2013-03-21 10:32:10: Listener for "2": stratum.btcguild.com:3333 21/03/2013 10:32:10, Setting new difficulty: 1

not sure if it was me messing with funky guiminer or what but glad you where able to see a task that you hadnt got to.

The main getwork server (btcguild.com:8332) was restarted to add back the X-Stratum redirect.

The DE server and merged mining servers will be taken down shortly in order to upgrade PoolServerJ to v2 blocks.

Ah shoot, thank you for pointing out that the updated PoolServerJ currently on getwork is not forwarding an X-Stratum header.  Time to fix that.

Port 8332 is the getwork port.  If you want to mine using stratum you probably should be using

Host (USA)    stratum.btcguild.com:3333
or
Host (EU)    eu-stratum.btcguild.com:3333

Deleted my poclbm.ini file and started over...I am able to mine now but am getting this in console still.

2013-03-21 10:00:56: Running command: poclbm.exe zif33rs_1:[email protected]:8332 --device=0 --platform=0 --verbose -r1 -v -w128 -f60
2013-03-21 10:00:56: Listener for "1" started
2013-03-21 10:00:57: Listener for "1": btcguild.com:8332 21/03/2013 10:00:57, checking for stratum...
2013-03-21 10:00:57: Listener for "1": btcguild.com:8332 21/03/2013 10:00:57, started OpenCL miner on platform 0, device 0 (Tahiti)
2013-03-21 10:00:58: Listener for "1": btcguild.com:8332 21/03/2013 10:00:58, using JSON-RPC (no stratum header)
2013-03-21 10:01:05: Running command: poclbm.exe zif33rs_2:[email protected]:8332 --device=2 --platform=0 --verbose -r1 -v -w128 -f60
2013-03-21 10:01:05: Listener for "2" started
2013-03-21 10:01:07: Listener for "2": btcguild.com:8332 21/03/2013 10:01:07, checking for stratum...
2013-03-21 10:01:07: Listener for "2": btcguild.com:8332 21/03/2013 10:01:07, started OpenCL miner on platform 0, device 2 (Devastator)
2013-03-21 10:01:07: Listener for "2": btcguild.com:8332 21/03/2013 10:01:07, using JSON-RPC (no stratum header)


seems like I had edited the .ini to read "stratum.btcguild.com:3333" and it worked just fine. not sure where the double port thing came from in the last .ini file.

You can enter two sets of port numbers??  I've never seen that before.

DDOS again?

rebooted my comp and tried to start mining...getting this.

2013-03-21 09:29:43: Running command: poclbm.exe zif33rs_1:[email protected]:3333:8332 --device=0 --platform=0 --verbose -r1 -v -w128 -f60
2013-03-21 09:29:43: Listener for "1" started
2013-03-21 09:29:44: Running command: poclbm.exe zif33rs_2:[email protected]:3333:8332 --device=2 --platform=0 --verbose -r1 -v -w128 -f60
2013-03-21 09:29:44: Listener for "2" started
2013-03-21 09:29:45: Listener for "2": stratum.btcguild.com:3333:8332 21/03/2013 09:29:45, checking for stratum...
2013-03-21 09:29:45: Listener for "2": stratum.btcguild.com:3333:8332 21/03/2013 09:29:45, no response to getwork, using as stratum
2013-03-21 09:29:45: Listener for "2": stratum.btcguild.com:3333:8332 21/03/2013 09:29:45, started OpenCL miner on platform 0, device 2 (Devastator)
2013-03-21 09:29:45: Listener for "2": Traceback (most recent call last):
2013-03-21 09:29:45: Listener for "2": File "poclbm.py", line 95, in
2013-03-21 09:29:45: Listener for "2": File "Switch.pyo", line 98, in loop
2013-03-21 09:29:45: Listener for "2": File "StratumSource.pyo", line 94, in loop
2013-03-21 09:29:45: Listener for "1": stratum.btcguild.com:3333:8332 21/03/2013 09:29:45, checking for stratum...
2013-03-21 09:29:45: Listener for "1": stratum.btcguild.com:3333:8332 21/03/2013 09:29:45, no response to getwork, using as stratum
2013-03-21 09:29:45: Listener for "1": stratum.btcguild.com:3333:8332 21/03/2013 09:29:45, started OpenCL miner on platform 0, device 0 (Tahiti)
2013-03-21 09:29:45: Listener for "1": Traceback (most recent call last):
2013-03-21 09:29:45: Listener for "1": File "poclbm.py", line 95, in
2013-03-21 09:29:45: Listener for "1": File "Switch.pyo", line 98, in loop
2013-03-21 09:29:45: Listener for "1": File "StratumSource.pyo", line 94, in loop


then workers show Kh\s thruput not up to the usual 700+ mh\s..is this the pool or something I f'ed up?

Thanks for trying maintening getwork ....

I still have my 3 (!) Pc's not working well with stratum, and running nicely with getwork...

as I have no time to find out what is the culprit (OS, Drivers, miner, you name it) ... I will stay on getwork so long I'm allowed to use it!

Getwork Saved(?)

Put in another late night, was able to get PSJ to generate valid v2 blocks on testnet.  I've restarted the primary getwork server, and we'll see if it carries over to livenet.  If it does, I'll be deployed it to the DE and mergedmining servers in the next 24 hours.

Please note that getwork is still a dead protocol that is planned for eventual removal once it has become a relatively insignificant portion of the pool hash rate.  It will also be removed with short notice if BTC Guild gets too close to 51% of the network.

Notice to Getwork Miners

Deepbit has switched to version 2 blocks, and Ozcoin has announced getwork will be closed this Saturday.  At this time, getwork may be forced to shut down on Sunday, March 24th.  I will not make BTC Guild the sole reason that this change does not take place, and I will also not risk a run of bad luck making the getwork servers unable to produce valid blocks due to the 95% supermajority being reached.  I am doing another run through the PoolServerJ code, and will be running it on testnet the next few days to see if I can buy some extra time.  However, please take this as notice that getwork WILL be removed, and plan on it being at the end of this weekend, even if it is able to limp on for a few extra days/weeks.

Update on Litecoin Status and Getwork.


Getwork will likely be removed by April 15th.  It may be removed earlier, but I will try to give at least one week notice if that is the case.  Merged mining will be killed off with getwork, since I still have no plans to work on implementing MM into my Stratum code.  NMC users will have a few weeks to easily withdraw NMC through the normal interface.  After that time it will be available through a special interface.

For those who do not want to move to Stratum, I am sorry but I cannot continue to support a protocol which was already terribly inefficient to begin with.  While there is still a fair amount of hash power on BTC Guild's getwork interface, we're also at a point where BTC Guild losing some of its hash power is a good thing.  The network has been converging on BTC Guild rapidly these last few weeks.  Hopefully BFL will deliver soon, because I strongly believe that the distribution of BFL-loyal miners does not favor BTC Guild as heavily as the distribution of Avalon customers.


At this time, I am not sure if Litecoin will be officially added.  After dealing with botnets this week, I'm cautious about adding Litecoin, which is about to become botnet heaven as a result of ASICs drastically raising the Bitcoin difficulty.  Once botnets realize Bitcoin isn't worth their time due to post-ASIC difficulty, you can be sure they're going to swarm Litecoin like locusts since it was designed as CPU (BOTNET) friendly.

Another busy day.  Just some small changes to the site today:

1) Pool speed chart is now a 90-day history and an area+spline chart rather than a simple line.
2) PPLNS Shifts increased to 10 million shares.  This setting should last for at least another month.  When it's time to change again, I will be working on an automatic shift size adjustment which uses pool speed.  Hopefully by the time that is needed, the pool speed will be somewhat stable.

PPLNS change once again has no noticeable affect on earnings when examined in a 24-hour period.  Larger shift sizes reduce per-shift variance, and help internally by preventing the database from needing to track too many shifts.  The only effect of the longer shift is that it takes more time for a share to be closed, so if you turned off your miners, it would take longer before your account balance stops increasing.  Current estimated mature time for a given share is ~8 hours at the current pool speed.

Servers should be back, 3 of the 5 US-based Stratum servers were under attack.

Pool getting DDoS'd by an annoyed botnet owner.  Working on bringing it back.

A lot of factors these last two weeks led to unstable stratum and also an error in the speed logs (that spike on your charts).  Has nothing to do with Stratum vs Getwork.  Although you will see more variance if that worker is running at higher difficulty, though 24-hour variance would be less than 1% (+ or -).




Stratum proxy running on a nettop (Acer AspireRevo) with a small FPGA farm hooked up to it.

On the subject of eliminating getwork...

I use the proxy