Topic: [~1000 GH/sec] BTC Guild - 0% Fee Pool, LP, SSL, Full Precision, and More - page 83. (Read 379078 times)

Just did some renaming and they didn't alphabetize themselves... was your alphabetizing a one time deal or should renames and new workers put themselves in the correct order?

Some people have their workers setup like this:

Computer_A-GPU_A
Computer_A-GPU_B
Computer_A-GPU_C
Computer_B-GPU_A
Computer_B-GPU_B
Computer_C-GPU_A
Computer_C-GPU_B


So, having everything alphabetized makes everything nice-n-tidy™

Why? I and Im sure a lot others had the workers in order and now they are in disorder.

I know you are just trying to improve the site, but maybe making it optional would be better?

1) is quite nice.  2) and 3) ... EXCELLENT!  That is better than anything any other pool does that I am aware of.  Later tonight, when I get the stats I am gathering, I will move back to your pool, payoff my ~-0.25BTC debt and then earn some more.  I am very happy indeed to see such an innovative approach to solving a common attack vector.  Further, it may keep some miners [i.e. Phoenix 1.48 .. jury is still out on 1.50 for me] from crashing outright in many cases.

As always, you are the best!

Thank you.

3 updates so far this morning:
1) Account workers alphabetized.
2) Getwork spammer detection script optimized (very quickly fixed NL1's issues).
3) Failed login timer - If your IP address has been failing logins, it will begin to enforce a timeout window before accepting new login credentials.  This timer grows exponentially after the third failed login.  It is entirely IP based, not account based, so a brute force attempt cannot circumvent the timer by successfully logging into a known account.

NICE!
Account workers are now alphabetized.   

I have an issue with phoenix on both when something changes with the pool [or several rapid long polling push go through].  POCLBM ALWAYS recovers for me.

EDIT:  I just noticed that they finally came out with phoenix 1.50 (never saw a 1.49).  Anyway, I tentatively updated the software [shared among machines making for easy upgrade ].  It is way to early to tell, but so far not a single stale; having said that, no stales in poclbm either, so your mileage may vary.  If it proves to be stable (no real statistical difference in stales compared to poclbm and no lockups or faults on long polling or idles) then I will move my other workers to Phoenix and away from poclbm.  Why?  Phoenix seems to maintain hashrate a little better than poclbm when the machine is in use (doesn't seem to matter what -f value I pass, but if I go much below 30 on my workstation that I use for work every day [9-13 hours straight] and play, the lower -f value causes a bit of desktop lag and hash rate still drops when moving windows around for instance [I have Aero turned on for this machine].  For whatever reason, Phoenix, doesn't give me that problem while maintaining a good balance (virtually no noticeable desktop lag, unless some site pops a YouTube video which results in first video and then a hard lock ... damn ATI/AMD drivers (Catalyst 11.6).  They have never been known to have all that stable of drivers for their latest series of hardware.

Here is what I use with Phoenix on my work station (heavy use).  Dual monitors with high resolution on both [not that cheap 1920x1080 crap they are selling now days, 2048x1152 and 1680x1050 respectively ... second is older and my side monitor, but both are 16x10 ratio which is much nicer than 16:9 if you do any development or documentation work IMHO].

MSI Radeon 6970 (essentially reference board) with core clock pushed just a little to 920MHz running at about 76-78C (have had to increase the fan ... I think it is time to blow dust and dander out ).

VECTORS BFI_INT FASTLOOP WORKSIZE=128 AGGRESSION=9 -k poclbm
VECTORS BFI_INT FASTLOOP WORKSIZE=128 AGGRESSION=9 -k poclbm

I got credit for 989.

Still showing as no credit for any work starting with block 989.  I guess you're welcome to the 15 or so 5830-hours as an unexpected donation -- hope it's the pool op who got the bonus instead of a random person. It's been a great pool until then.

Also confirming I'm getting way fewer connection/stale issues at deepbit with the phoenix miner.  Very short and very occasional miner idles, and actual credit for work done. 

I am sure you have tried it, but Diablo Miner is my favorite. Very active developer and with the vectors and -f1 I get the same hash rate as phoenix.

Yeh using phoenix, I have no issue with phoenix on deepbit ?

Using Phoenix?  It isn't stable with long polling.    Yesterday, three blocks were solved in rapid succession leading to three long polling pushes in very close proximity to each other; for me, two in a row with no submitted shares between the two.  Went for another share or two if I remember correctly and then I noticed the red color of the worker on the site and went to take a look.  Found it sitting there, looking as if it was running [showed full hash rate, history, accepted/rejected shares], but GPU was at 0% utilization and I noticed that phoenix couldn't even be stopped via the normal Ctrl-C ... bye.  Killing the shell window stopped the process easily enough and I restarted.  Problem is that Phoenix is always giving me trouble [active development die on phoenix or what?].

Hey, somebody made poclbm with the phatk kernel, maybe this is something for you xD
http://forum.bitcoin.org/index.php?topic=19169.0

miner idle message don't seem to be working.  my rig just went down for an hour due to a power outage and no email came. 

next to the miner idle setting is a warning saying that my donation for the previous round was under 2%, but my donation is set for 2.5% and the unconfirmed reward is 0, with everything in confirmed, so that is working.

mmm starting to think long polling isnt working all that well(or at all?) on btcguild.

I am consistently >4% stales with btcguild and my deepbit is consistently <2%

lol I had to run downstairs and double check my machines after I posted ;o)

In the last 2 hours or so I've gotten 11 miner is idle and I think it was 6 prob comm with rpc etc.. I've personally had MUCH worse with other pools.

Not trying to convince anyone of anything, just for me 11 is a lot less than I've previously had, but unlike the previous poster I'm not seeing my gpu cycle excessively or my temps drop noticeably. To each his/her own ;o)

I'm connected to the servers with the lowest ping from where I'm at.  Still getting HORRIBLE connection issues.  It's pretty much sunk my mining at BTC for tonight I think.  ;-(

A lotta idles again

For you all having connection problems, and Really I imagine most of you have done this, but when more pools servers came online I did like 20 pings or so to each server. For me a certain server was dropping every 4th packet, so I obviously didn't use it. I then picked the server with the lowest latency ~60-90 was my lowest. I'm now getting very few stales. One of my miners has crept up a little more than I'd like, but I've also been playing with longer askrates.

Have you guys tried doing the same?

eleuthria - I just wanted to say that I jumped pools temporarily.  I have a negative balance, (not much ~0.25BTC), but I switched due to some idle issues [not long, but long enough and frequent enough to noticeably allow GPU temperature changes].  I will be back shortly, either in a few days [Monday at the latest] or as soon as I have similar problems again [I often find that Phoenix has either dropped to 0MH/s or it just stops as if it is mining showing a hash rate as normal(won't respond to Ctrl-C), long polling in quick succession from the pool I switched too often gives Phoenix on one of the machines fits and occasionally takes out both machines I use it on].  I get the occasional issue with Phoenix on BTC Guild as well, but it is far rarer, which is one among many reasons I prefer your pool.  POCLBM miner is my preference and what I use with my 6970, but the phatk kernel in phoenix gives an ~3% increase in hash rate on my 5850s at the expense of an increase of 1% stales (so net is good if the miners stay on line).  For the record, since Phoenix is never fixed, I hate Phoenix ... love hate ... I like the output a little better, but it is the most unreliable miner  ... I wish they would fix it, but no news is bad news. 

More information perhaps than I needed, but I wanted to let you know that I am not jumping out of the negative balance [that would be unethical, and it simply isn't me].  I will mop up the debt soon.

Thank you for the excellent pool!