[CLOSED] BTC Guild - Pays TxFees+NMC, Stratum, VarDiff, Private Servers

[dhenson]

Anyone know what 'other' errors are in the btcguild dashboard? I'm down to 400GH/s and about 1/4th of the total shares are ending up in the OTHER error column.

[-Redacted-]

One guess would be things like shares submitted at the wrong difficulty.

[eleuthria]

Anyone know what 'other' errors are in the btcguild dashboard? I'm down to 400GH/s and about 1/4th of the total shares are ending up in the OTHER error column.

What miner are you using, it's SEVERELY broken.  You're submitting 500k+ shares that didn't come from BTC Guild.  Your 500k+ "Other" shares are Unknown Work, meaning you're returning results to work that was never given to you, either because your miner is spitting out corrupted data/unpadded results, or submitting work from a different pool.

[dhenson]

Anyone know what 'other' errors are in the btcguild dashboard? I'm down to 400GH/s and about 1/4th of the total shares are ending up in the OTHER error column.

What miner are you using, it's SEVERELY broken.  You're submitting 500k+ shares that didn't come from BTC Guild.  Your 500k+ "Other" shares are Unknown Work, meaning you're returning results to work that was never given to you, either because your miner is spitting out corrupted data/unpadded results, or submitting work from a different pool.

It's a hosted KNC Jupiter.  That doesn't surprise me as it doesn't seem to keep it's settings and randomly switches pools on me.  Thank you for the info.

[gnar1ta$]

Anyone know what 'other' errors are in the btcguild dashboard? I'm down to 400GH/s and about 1/4th of the total shares are ending up in the OTHER error column.

What miner are you using, it's SEVERELY broken.  You're submitting 500k+ shares that didn't come from BTC Guild.  Your 500k+ "Other" shares are Unknown Work, meaning you're returning results to work that was never given to you, either because your miner is spitting out corrupted data/unpadded results, or submitting work from a different pool.

This may actually just be a "feature" of the crappy KNC design.  Here is mine:

Code:

479.51 GH/s	42,480k (99.34%)	59392 / 125k / 223k

[eleuthria]

This may actually just be a "feature" of the crappy KNC design.  Here is mine:

Code:

479.51 GH/s	42,480k (99.34%)	59392 / 125k / 223k

Strange, considering I know quite a few KNC units that mine on BTC Guild since day 1, and have only just heard this now.  Are you two running the latest firmware?  Was it doing this *before* the latest firmware?  Multiple pools configured?

[dhenson]

This may actually just be a "feature" of the crappy KNC design.  Here is mine:

Code:

479.51 GH/s	42,480k (99.34%)	59392 / 125k / 223k

Strange, considering I know quite a few KNC units that mine on BTC Guild since day 1, and have only just heard this now.  Are you two running the latest firmware?  Was it doing this *before* the latest firmware?  Multiple pools configured?

I'm a hosted customer so I don't have access to know exactly what firmware I am running.  However, my hashrate was hovering around 120GH for the first few days and once they announced .95 I started seeing ~ 420GH so I assume that is what I am running.

Unfortunately, at this time, I don't have SSH access to do any troubleshooting, but the general feeling over in the KNC thread is that I have two copies of cgminer running on my box.

Their interface only lets you do one thing, and that is set your pool information and 'save'.  You can add as many pool configs as you want and my initial assumption is that they would act as backup pools.  However, my unit is doing two odd things in regards to the pools.

1) after saving my config, the unit appears to hash at the new pool only to revert back to the previous pool after a few minutes (supporting the two cgminer process theory)
2) it appears that work is being accepted from the previous pool and then being submitted to the new pool causing the 'other' errors.

I've been having these issues since the beginning. (hosted issue only?)

[Taugeran]

ok idk where to put this so @eleuthria feel free to delete but in under 30 days we have added 1.25 Quadrillion Hashes per second (Ph/s).
in 30 days we have increased the daily average hash rate by 125% (according to block chain)

it may be b/c I've not had my coffee yet but i cannot seem to believe it

[gnar1ta$]

This may actually just be a "feature" of the crappy KNC design.  Here is mine:

Code:

479.51 GH/s	42,480k (99.34%)	59392 / 125k / 223k

Strange, considering I know quite a few KNC units that mine on BTC Guild since day 1, and have only just heard this now.  Are you two running the latest firmware?  Was it doing this *before* the latest firmware?  Multiple pools configured?

Others may not be looking, I wasn't.  I didn't even realize I had so much other shares until dhenson brought it up.  My hash rate goes between 470 and 530 and my accepted % is good, and I'm getting paid, so is it even an issue?  I have the latest firmware.  Don't know about previous versions.

[eleuthria]

Others may not be looking, I wasn't.  I didn't even realize I had so much other shares until dhenson brought it up.  My hash rate goes between 470 and 530 and my accepted % is good, and I'm getting paid, so is it even an issue?  I have the latest firmware.  Don't know about previous versions.

I guess that is another factor.  If the units internal reporting and BTC Gulid's dashboard are fairly close after an hour, it may be that these rejects aren't even "wasted" work.  I know my old ztex FPGAs used to have a problem with duplicate shares, anywhere from 8 to 10%.  However, they weren't doing duplicate work, it was just a bug in the firmware not clearing out the "winning nonce" part of memory when it submitted the share, so it would resubmit it each time it finished a full nonce range without a new result [pretty sure that's what it was at least].  The dashboard/payouts were matching the speed reported by the mining software.

[gnar1ta$]

@dhenson

What are your pool settings, PPS PPNLS, and var diff or static diff? And what pool address are you using to connect? (don't post the private address if you're using that)

[dhenson]

@dhenson

What are your pool settings, PPS PPNLS, and var diff or static diff? And what pool address are you using to connect? (don't post the private address if you're using that)

I'm mining to btcguild ppnls var diff, to the public pool address stratum+tcp://stratum.btcguild.com:3333

[gnar1ta$]

@dhenson

What are your pool settings, PPS PPNLS, and var diff or static diff? And what pool address are you using to connect? (don't post the private address if you're using that)

I'm mining to btcguild ppnls var diff, to the public pool address stratum+tcp://stratum.btcguild.com:3333

Yeah change your diff from default to 256.  I know this is a short sample, but here is around 30 mins of var diff vs min:

default diff

Code:

452.95 GH/s   235916 (99.74%)   256 / 768 / 364

min 256

Code:

 470.14 GH/s   203008 (100.00%)   0 / 0 / 0

They run the ridiculously slow flushwork not just on new blocks but also on new diff work - when the diff changes I get a lot of rejected shares on cgminer.  I suspect this would be compounded by fast block finds, like on p2pool, and on miner restarts when the work starts at diff 1.

[eleuthria]

Just FYI:  There is no "static diff" on BTC Guild.  There is a *minimum difficulty* setting, which starts your miner off at your minimum and won't adjust it below that.  Vardiff will always adjust you *higher* if your hardware is pushing out too many shares.

[gnar1ta$]

Thanks, I fixed it.  But either way, min or static, make it high enough that it won't change seems to be the solution.  Unless KNC decides to fix it on their end.

[Trongersoll]

None of the charts on the pool website seem to be working for me. 

i'm running Windows Internet Explorer 8 on Windows XP.

[eleuthria]

None of the charts on the pool website seem to be working for me. 

i'm running Windows Internet Explorer 8 on Windows XP.

Should be fixed now.  There was a random issue with certain JS failing to load due to some optimizations I added to reduce overall page load time.  Recently added a more complete fix to the issue, so on a refresh you should have all the charts working again.

[-Redacted-]

Can't reach the webpage for some reason. 

[Trongersoll]

Can't reach the webpage for some reason. 

Yeah, what he said. Are we under attack again? Do we need to man the guns? Are the shields up?

[-Redacted-]

Back now.