[OLD] Eligius: ASIC, no registration, no fee CPPSRB BTC + 105% PPS NMC, 877 # - page 48.

helmax

sr. member

Activity: 440

Merit: 250

Quote from: wizkid057 on December 16, 2013, 12:58:38 AM

Anyone with an affected KnC device is welcome to switch to and try a copy of the dev server I put on one of the main IPs.

v2.srv.eligius.st:12234

Let me know if this helps/hurts/etc please. I'll check back tomorrow to perhaps bug someone for remote access again to test.

-wk

works good drop to 3.2% HW but can drop more
others pools only have 2.0% nov jupiter

btcspender

member

Activity: 109

Merit: 10

Quote from: wizkid057 on December 16, 2013, 12:58:38 AM

Anyone with an affected KnC device is welcome to switch to and try a copy of the dev server I put on one of the main IPs.

v2.srv.eligius.st:12234

Let me know if this helps/hurts/etc please. I'll check back tomorrow to perhaps bug someone for remote access again to test.

-wk

Yep the dev server is working great low 1.4% - 2% rate on oct jup and 2% - 2.5% on nov jup & no cores disabling on nov jup so whatever the difference of the dev sever it has fixed the problem for me

davebodger

full member

Activity: 154

Merit: 100

Quote from: wizkid057 on December 16, 2013, 12:58:38 AM

Anyone with an affected KnC device is welcome to switch to and try a copy of the dev server I put on one of the main IPs.

v2.srv.eligius.st:12234

Let me know if this helps/hurts/etc please. I'll check back tomorrow to perhaps bug someone for remote access again to test.

-wk

Well my HW% was 6.8% and I swapped pool (to dev) and it's dropped to 3.3% and is still falling. I'm pulling a few watts more from the wall, which is good I think. I have not spotted any cores disabling either.
Only tested over 30 mins so far, will report back again when I get home from work, but looking good so far.

Only difference I've spotted so far is on dev I'm getting diff varying between 128/256 wheras on normal pool server it varies between 512/1024.

Regards.

Dave.

wizkid057

legendary

Activity: 1223

Merit: 1006

Anyone with an affected KnC device is welcome to switch to and try a copy of the dev server I put on one of the main IPs.

v2.srv.eligius.st:12234

Let me know if this helps/hurts/etc please. I'll check back tomorrow to perhaps bug someone for remote access again to test.

-wk

wizkid057

legendary

Activity: 1223

Merit: 1006

Quote from: lenny_ on December 15, 2013, 11:28:45 AM

Quote from: wizkid057 on December 15, 2013, 11:16:56 AM

Quote from: lenny_ on December 15, 2013, 11:13:11 AM

I am confirming, that main Eligius server still have issues November Jupiters HW ratio.

Dev-server (IP given by wizkid057), results:
Device Hardware%=2.0833, WU 9158
https://docs.google.com/file/d/0B2z0CgjT8HdwWUZ2cDc5VXp2TFE

Main server, results:
Device Hardware%=11.0150 ,WU 8289
https://docs.google.com/file/d/0B2z0CgjT8HdwMEtISk5xLThjaVU
(Please notice lots of cores disabled due to increased HW ratio)

I am switching Jups back to dev-server, looks like main server still have this issue.

Not sure how that makes any sense at this point. The main server is running the exact same code now.

I'll look into it further.

-wk

Thank you. Hope you'll find it.

I switched back to dev-server and immediately got better results (live):

Code:

Avg. Hash Rate 668 Gh/s
WU 9132
Difficulty Accepted 108798
Device Hardware%=1.9638

I've been checking into this for a while now and there seems to be no real differences between the dev server and the main server now besides the IP/port... I'm very confused as to why there would be any difference in performance.

Tomorrow I will have to bother some of the people who've been helping me for my remote-access KnC time to see if I can figure this out, again.

-wk

wizkid057

legendary

Activity: 1223

Merit: 1006

Quote from: zuqka on December 15, 2013, 05:08:55 PM

Quote from: wizkid057 on December 15, 2013, 04:49:14 PM

Quote from: zuqka on December 15, 2013, 04:30:11 PM

Well for the last 12 hrs my 180Ghs (3xBitburner Furys & 1 Bitfury) showed 140Ghs. Checked the equipment and everything was ok. Watching it for the last 30min and I it looks like it started to show the correct hashing rate again. Since we got less coins because of false values will there be any reimbursement?

False values? There is no pool side issue at work here. The only known issue right now is that the cgminer + knc driver have some kind of issue.

-wk

Sorry how would you call it? From 03:00 until 18:22 today I was getting 30Ghs less shown on your pool then my rigs were hashing! If you dont care to check thats your prob.

I have no way to even check anything at all if you provide zero information to utilize.

lenny_

legendary

Activity: 1036

Merit: 1000

DARKNETMARKETS.COM

Quote from: zuqka on December 15, 2013, 05:08:55 PM

Quote from: wizkid057 on December 15, 2013, 04:49:14 PM

Quote from: zuqka on December 15, 2013, 04:30:11 PM

Well for the last 12 hrs my 180Ghs (3xBitburner Furys & 1 Bitfury) showed 140Ghs. Checked the equipment and everything was ok. Watching it for the last 30min and I it looks like it started to show the correct hashing rate again. Since we got less coins because of false values will there be any reimbursement?

False values? There is no pool side issue at work here. The only known issue right now is that the cgminer + knc driver have some kind of issue.

-wk

Sorry how would you call it? From 03:00 until 18:22 today I was getting 30Ghs less shown on your pool then my rigs were hashing! If you dont care to check thats your prob.

If you not happy please feel free to leave. Don't even waste developer's time without providing logs, stats, description of situation etc. Goodbye.

zuqka

full member

Activity: 140

Merit: 100

Quote from: wizkid057 on December 15, 2013, 04:49:14 PM

Quote from: zuqka on December 15, 2013, 04:30:11 PM

Well for the last 12 hrs my 180Ghs (3xBitburner Furys & 1 Bitfury) showed 140Ghs. Checked the equipment and everything was ok. Watching it for the last 30min and I it looks like it started to show the correct hashing rate again. Since we got less coins because of false values will there be any reimbursement?

False values? There is no pool side issue at work here. The only known issue right now is that the cgminer + knc driver have some kind of issue.

-wk

Sorry how would you call it? From 03:00 until 18:22 today I was getting 30Ghs less shown on your pool then my rigs were hashing! If you dont care to check thats your prob.

wizkid057

legendary

Activity: 1223

Merit: 1006

Quote from: zuqka on December 15, 2013, 04:30:11 PM

Well for the last 12 hrs my 180Ghs (3xBitburner Furys & 1 Bitfury) showed 140Ghs. Checked the equipment and everything was ok. Watching it for the last 30min and I it looks like it started to show the correct hashing rate again. Since we got less coins because of false values will there be any reimbursement?

False values? There is no pool side issue at work here. The only known issue right now is that the cgminer + knc driver have some kind of issue.

-wk

zuqka

full member

Activity: 140

Merit: 100

Well for the last 12 hrs my 180Ghs (3xBitburner Furys & 1 Bitfury) showed 140Ghs. Checked the equipment and everything was ok. Watching it for the last 30min and I it looks like it started to show the correct hashing rate again. Since we got less coins because of false values will there be any reimbursement?

RHA

sr. member

Activity: 392

Merit: 250

I confirm. My experience with it is near identical like lenny_'s.
What's more, October Jupiter has around 4% HW with main servers, while 2.4% HW with dev machine and 1.9% HW with bitminter.

lenny_

legendary

Activity: 1036

Merit: 1000

DARKNETMARKETS.COM

Quote from: Luke-Jr on December 15, 2013, 12:34:10 PM

I don't get why people still tolerate cgminer and all its many bugs.

It's on KNC firmware by default, I have to stick with it for now, I tried bfgminer from BertMod, but it didn't work very well, throwed out lots of errors and started mining with 20 GH/s or so.

Luke-Jr

legendary

Activity: 2576

Merit: 1186

I don't get why people still tolerate cgminer and all its many bugs.

lenny_

legendary

Activity: 1036

Merit: 1000

DARKNETMARKETS.COM

Quote from: wizkid057 on December 15, 2013, 11:16:56 AM

Quote from: lenny_ on December 15, 2013, 11:13:11 AM

I am confirming, that main Eligius server still have issues November Jupiters HW ratio.

Dev-server (IP given by wizkid057), results:
Device Hardware%=2.0833, WU 9158
https://docs.google.com/file/d/0B2z0CgjT8HdwWUZ2cDc5VXp2TFE

Main server, results:
Device Hardware%=11.0150 ,WU 8289
https://docs.google.com/file/d/0B2z0CgjT8HdwMEtISk5xLThjaVU
(Please notice lots of cores disabled due to increased HW ratio)

I am switching Jups back to dev-server, looks like main server still have this issue.

Not sure how that makes any sense at this point. The main server is running the exact same code now.

I'll look into it further.

-wk

Thank you. Hope you'll find it.

I switched back to dev-server and immediately got better results (live):

Code:

Avg. Hash Rate	668 Gh/s
WU	9132
Difficulty Accepted	108798
Device Hardware%=1.9638

wizkid057

legendary

Activity: 1223

Merit: 1006

Quote from: lenny_ on December 15, 2013, 11:13:11 AM

I am confirming, that main Eligius server still have issues November Jupiters HW ratio.

Dev-server (IP given by wizkid057), results:
Device Hardware%=2.0833, WU 9158
https://docs.google.com/file/d/0B2z0CgjT8HdwWUZ2cDc5VXp2TFE

Main server, results:
Device Hardware%=11.0150 ,WU 8289
https://docs.google.com/file/d/0B2z0CgjT8HdwMEtISk5xLThjaVU
(Please notice lots of cores disabled due to increased HW ratio)

I am switching Jups back to dev-server, looks like main server still have this issue.

Not sure how that makes any sense at this point. The main server is running the exact same code now.

I'll look into it further.

-wk

lenny_

legendary

Activity: 1036

Merit: 1000

DARKNETMARKETS.COM

I am confirming, that main Eligius server still have issues November Jupiters HW ratio.

Dev-server (IP given by wizkid057), results:
Device Hardware%=2.0833, WU 9158
https://docs.google.com/file/d/0B2z0CgjT8HdwWUZ2cDc5VXp2TFE

Main server, results:
Device Hardware%=11.0150 ,WU 8289
https://docs.google.com/file/d/0B2z0CgjT8HdwMEtISk5xLThjaVU
(Please notice lots of cores disabled due to increased HW ratio)

I am switching Jups back to dev-server, looks like main server still have this issue.

wizkid057

legendary

Activity: 1223

Merit: 1006

Quote from: btcspender on December 15, 2013, 08:20:56 AM

Quote from: wizkid057 on December 14, 2013, 10:57:20 PM

Still working on updates to the servers.

I applied the server-side patch to work around the KNC cgminer driver issue that should help KNC users. I'm already seeing improvements across the board for KNC folks.

-wk

I moved my oct jup to ghash a fair few hours ago as hw errors were at about 4% from about 1.8% which was strange as it was nov jup that was having the hw error issue I dont know how long the oct jup errors were at 4% I hadn't checked in a day or 2. I just moved them both back to eligius now and within 5 - 10 mins oct jup climbs to 4% and nov jup to 10% hw errors and cores have started to turn off on nov jup but no cores off on oct jup. The only other pool I've used lately besides eligius is ghash and both jups stabilise at about 1.8% hw errors nov jup sometimes is at around 2% or a bit over.

Cant wait to get my jups back to eligius ghash payouts are woeful compared to eligius

edit:
jups now at 5% and 12% back to ghash while I sleep I think, I guess I'll see when I wake up if I'm the only one still with knc problems.

You're so far the only person to report any issue since I made the server-side changes to the work update interval.

Looking at several known KnC miner users' graphs I see notable improvements. Also speaking with folks on IRC I'm told by folks who were having the issue that they are not any longer.

I'd like to investigate further however if you are still having issues.

Edit: I assume you're using Eligius stratum, also.

-wk

helmax

sr. member

Activity: 440

Merit: 250

thanks wizkid057 we need fix for hw kncminer
and server europe

btcspender

member

Activity: 109

Merit: 10

Quote from: wizkid057 on December 14, 2013, 10:57:20 PM

Still working on updates to the servers.

I applied the server-side patch to work around the KNC cgminer driver issue that should help KNC users. I'm already seeing improvements across the board for KNC folks.

-wk

I moved my oct jup to ghash a fair few hours ago as hw errors were at about 4% from about 1.8% which was strange as it was nov jup that was having the hw error issue I dont know how long the oct jup errors were at 4% I hadn't checked in a day or 2. I just moved them both back to eligius now and within 5 - 10 mins oct jup climbs to 4% and nov jup to 10% hw errors and cores have started to turn off on nov jup but no cores off on oct jup. The only other pool I've used lately besides eligius is ghash and both jups stabilise at about 1.8% hw errors nov jup sometimes is at around 2% or a bit over.

Cant wait to get my jups back to eligius ghash payouts are woeful compared to eligius

edit:
jups now at 5% and 12% back to ghash while I sleep I think, I guess I'll see when I wake up if I'm the only one still with knc problems.

wizkid057

legendary

Activity: 1223

Merit: 1006

Still working on updates to the servers.

I applied the server-side patch to work around the KNC cgminer driver issue that should help KNC users. I'm already seeing improvements across the board for KNC folks.

-wk

Topic: [OLD] Eligius: ASIC, no registration, no fee CPPSRB BTC + 105% PPS NMC, 877 # - page 48. (Read 458370 times)