Status as of Tuesday, July 10, 2018

Discussion of development releases of Prohashing.
User avatar
Steve Sokolowski
Posts: 3323
Joined: Wed Aug 27, 2014 3:27 pm
Location: State College, PA
Contact:

Status as of Tuesday, July 10, 2018

Postby Steve Sokolowski » Tue Jul 10, 2018 8:33 am

Good morning!

  • Yesterday, Chris made some share corrections for x11 and SHA-256 miners, for which we apologize. There was a problem that caused some customers to be overcredited when they were assigned coins with low difficulties. I resolved the issue and Chris deployed the changes. Fortunately, while I thought we would be inundated with tickets about this problem overnight, there hasn't been a single complaint.
  • Vance is hard at work with ETH mining. He already got ETH shares to be accepted and is now working on other ethash coins.
  • My goal today will be to get down to a lot of the more difficult support tickets, now that the system is very stable. There are currently 71 tickets outstanding, down from 95 two days ago. Many of those tickets were moved from "waiting for support" to "waiting for customer" last night after Chris put 4 hours into resolving tickets.
  • A major change was released this morning, at around 2:43am EDT. The mining server has been changed to reassign miners to new coins much less aggressively than it did before. We expect profitability to decline, but for hashrate to increase and offset the profitability loss. Static coin miners and solo miners are not affected by the change. After the system has run in this mode for a full day, which means about 40 hours from now, we will solicit feedback from customers on whether the change should be continued.
  • There was another subtle bug that was addressed yesterday. When the system was released, the hashrate of miners was low, so the initial setting of banning miners who submit 2500 rejected shares, with 25 shares forgiven every 300 seconds, was reasonable. Now, it was possible for a person to hit the 2500 share threshold very quickly. This issue was pointed out by a high hashrate miner. I changed the rejected share ban threshold so that 35,000 shares are now required to be banned, and so that 7,500 shares are forgiven every 300 seconds. This way, it will now take a lot of rejected shares to be banned and it is less likely for miners to be disconnected simply for having a high hashrate, where 1% rejected of 1 million shares would have resulted in a ban.
jeffms2003
Posts: 43
Joined: Fri May 11, 2018 11:40 am

Re: Status as of Tuesday, July 10, 2018

Postby jeffms2003 » Wed Jul 11, 2018 9:01 am

Well I moved everything over to Prohashing again over night and still getting low hash rates and about the same profitability as single coin pools because of the low hash rates being calculated. All the other pools give me full credit. I will continue to test Prohashing, but right now something just doesn't reconcile to the claims made in terms of profitability. Wonder if all the restarts are hard on the equipment too.
User avatar
Steve Sokolowski
Posts: 3323
Joined: Wed Aug 27, 2014 3:27 pm
Location: State College, PA
Contact:

Re: Status as of Tuesday, July 10, 2018

Postby Steve Sokolowski » Wed Jul 11, 2018 9:28 am

jeffms2003 wrote:Well I moved everything over to Prohashing again over night and still getting low hash rates and about the same profitability as single coin pools because of the low hash rates being calculated. All the other pools give me full credit. I will continue to test Prohashing, but right now something just doesn't reconcile to the claims made in terms of profitability. Wonder if all the restarts are hard on the equipment too.


We made a significant change to the way that coin switches occur two nights ago. If work restarts were responsible for reduced hashrates, then the new changes should have improved or fully resolved that issue. Since that doesn't appear to be the case, then, at least for you, there must be a different issue causing the problem. Are you sure there hasn't been any improvement at all?
jeffms2003
Posts: 43
Joined: Fri May 11, 2018 11:40 am

Re: Status as of Tuesday, July 10, 2018

Postby jeffms2003 » Wed Jul 11, 2018 9:44 am

Steve Sokolowski wrote:
jeffms2003 wrote:Well I moved everything over to Prohashing again over night and still getting low hash rates and about the same profitability as single coin pools because of the low hash rates being calculated. All the other pools give me full credit. I will continue to test Prohashing, but right now something just doesn't reconcile to the claims made in terms of profitability. Wonder if all the restarts are hard on the equipment too.


We made a significant change to the way that coin switches occur two nights ago. If work restarts were responsible for reduced hashrates, then the new changes should have improved or fully resolved that issue. Since that doesn't appear to be the case, then, at least for you, there must be a different issue causing the problem. Are you sure there hasn't been any improvement at all?


Scrypt is pretty close to actual, so that is an improvement, but still a little low. Equihash got way worse as I'm hashing at 14k sols and all morning the prohashing calculations have shown 11k (Equihash was close the last time I had it on here). SHA is about a 2% improvement, but still 4-5% low. When I extrapolate my revenue over a 7 hour period it is in line with if not slightly lower than what I would get at separate pools. I don't know what the issue is. Maybe my machines just don't like the way your servers do things or something. I don't think these things were made to do constant variable coin mining.
User avatar
Steve Sokolowski
Posts: 3323
Joined: Wed Aug 27, 2014 3:27 pm
Location: State College, PA
Contact:

Re: Status as of Tuesday, July 10, 2018

Postby Steve Sokolowski » Wed Jul 11, 2018 10:49 am

jeffms2003 wrote:
Steve Sokolowski wrote:
jeffms2003 wrote:Well I moved everything over to Prohashing again over night and still getting low hash rates and about the same profitability as single coin pools because of the low hash rates being calculated. All the other pools give me full credit. I will continue to test Prohashing, but right now something just doesn't reconcile to the claims made in terms of profitability. Wonder if all the restarts are hard on the equipment too.


We made a significant change to the way that coin switches occur two nights ago. If work restarts were responsible for reduced hashrates, then the new changes should have improved or fully resolved that issue. Since that doesn't appear to be the case, then, at least for you, there must be a different issue causing the problem. Are you sure there hasn't been any improvement at all?


Scrypt is pretty close to actual, so that is an improvement, but still a little low. Equihash got way worse as I'm hashing at 14k sols and all morning the prohashing calculations have shown 11k (Equihash was close the last time I had it on here). SHA is about a 2% improvement, but still 4-5% low. When I extrapolate my revenue over a 7 hour period it is in line with if not slightly lower than what I would get at separate pools. I don't know what the issue is. Maybe my machines just don't like the way your servers do things or something. I don't think these things were made to do constant variable coin mining.


If scrypt is at actual, then that means that you should be earning about 10% over LTC, so that's great. I'm also heartened to hear about the SHA-256 improvement. It's interesting to see that the hashrate charts show that hashrate rose over the past few days, despite there being little change in the number of connected miners.

Has anyone else had this equihash problem, or is it just you?
jeffms2003
Posts: 43
Joined: Fri May 11, 2018 11:40 am

Re: Status as of Tuesday, July 10, 2018

Postby jeffms2003 » Wed Jul 11, 2018 10:53 am

Steve Sokolowski wrote:
jeffms2003 wrote:
Steve Sokolowski wrote:
We made a significant change to the way that coin switches occur two nights ago. If work restarts were responsible for reduced hashrates, then the new changes should have improved or fully resolved that issue. Since that doesn't appear to be the case, then, at least for you, there must be a different issue causing the problem. Are you sure there hasn't been any improvement at all?


Scrypt is pretty close to actual, so that is an improvement, but still a little low. Equihash got way worse as I'm hashing at 14k sols and all morning the prohashing calculations have shown 11k (Equihash was close the last time I had it on here). SHA is about a 2% improvement, but still 4-5% low. When I extrapolate my revenue over a 7 hour period it is in line with if not slightly lower than what I would get at separate pools. I don't know what the issue is. Maybe my machines just don't like the way your servers do things or something. I don't think these things were made to do constant variable coin mining.


If scrypt is at actual, then that means that you should be earning about 10% over LTC, so that's great. I'm also heartened to hear about the SHA-256 improvement. It's interesting to see that the hashrate charts show that hashrate rose over the past few days, despite there being little change in the number of connected miners.

Has anyone else had this equihash problem, or is it just you?


I can assure you I'm not making 10% over litecoinpool.org, but maybe based on your calculations of LTC. As someone that is trying to make sure I make the most money for my investment, I'd rather know how your earnings per mh/s/day compares to something like that of litecoinpool.org because my decision will be made based on that, not what your calculation of LTC is. There is more value in seeing your numbers compared to top competitors, not internally set calculations.
User avatar
Steve Sokolowski
Posts: 3323
Joined: Wed Aug 27, 2014 3:27 pm
Location: State College, PA
Contact:

Re: Status as of Tuesday, July 10, 2018

Postby Steve Sokolowski » Wed Jul 11, 2018 3:01 pm

jeffms2003 wrote:
Steve Sokolowski wrote:
jeffms2003 wrote:
Scrypt is pretty close to actual, so that is an improvement, but still a little low. Equihash got way worse as I'm hashing at 14k sols and all morning the prohashing calculations have shown 11k (Equihash was close the last time I had it on here). SHA is about a 2% improvement, but still 4-5% low. When I extrapolate my revenue over a 7 hour period it is in line with if not slightly lower than what I would get at separate pools. I don't know what the issue is. Maybe my machines just don't like the way your servers do things or something. I don't think these things were made to do constant variable coin mining.


If scrypt is at actual, then that means that you should be earning about 10% over LTC, so that's great. I'm also heartened to hear about the SHA-256 improvement. It's interesting to see that the hashrate charts show that hashrate rose over the past few days, despite there being little change in the number of connected miners.

Has anyone else had this equihash problem, or is it just you?


I can assure you I'm not making 10% over litecoinpool.org, but maybe based on your calculations of LTC. As someone that is trying to make sure I make the most money for my investment, I'd rather know how your earnings per mh/s/day compares to something like that of litecoinpool.org because my decision will be made based on that, not what your calculation of LTC is. There is more value in seeing your numbers compared to top competitors, not internally set calculations.


That's what I'm trying to figure out. Since I don't have the same setup as you, it's difficult for me to determine exactly what the difference is between litecoinpool.org and our earnings.

I described the way we calculate the theoretical maximum LTC profitability. The estimate includes a 2.5% fee, which is actually even lower than any pay-per-share pool I've seen. The only way that litecoinpool could do better than that is to eliminate the fee, or to go further and actually lose money.

Could you tell me what the exact percentage difference is?
Tomas_Novak
Posts: 28
Joined: Wed Jun 13, 2018 4:14 am

Re: Status as of Tuesday, July 10, 2018

Postby Tomas_Novak » Wed Jul 11, 2018 3:55 pm

Today from 16:21 GMT I have sha 150 Th diameter up to now. It is 20% less than the last two days.
jeffms2003
Posts: 43
Joined: Fri May 11, 2018 11:40 am

Re: Status as of Tuesday, July 10, 2018

Postby jeffms2003 » Wed Jul 11, 2018 9:51 pm

Steve Sokolowski wrote:
jeffms2003 wrote:
Steve Sokolowski wrote:
If scrypt is at actual, then that means that you should be earning about 10% over LTC, so that's great. I'm also heartened to hear about the SHA-256 improvement. It's interesting to see that the hashrate charts show that hashrate rose over the past few days, despite there being little change in the number of connected miners.

Has anyone else had this equihash problem, or is it just you?


I can assure you I'm not making 10% over litecoinpool.org, but maybe based on your calculations of LTC. As someone that is trying to make sure I make the most money for my investment, I'd rather know how your earnings per mh/s/day compares to something like that of litecoinpool.org because my decision will be made based on that, not what your calculation of LTC is. There is more value in seeing your numbers compared to top competitors, not internally set calculations.


That's what I'm trying to figure out. Since I don't have the same setup as you, it's difficult for me to determine exactly what the difference is between litecoinpool.org and our earnings.

I described the way we calculate the theoretical maximum LTC profitability. The estimate includes a 2.5% fee, which is actually even lower than any pay-per-share pool I've seen. The only way that litecoinpool could do better than that is to eliminate the fee, or to go further and actually lose money.

Could you tell me what the exact percentage difference is?


I have 4 L3++ running, so I'm willing to run 2 on here and 2 on there for the same 24 hour period and we'll see how close they come. There is pretty good detail on how they run and the reasoning behind how/why they do the 101% PPS payouts. If I remember correctly they keep the merged mining profits to cover costs. I'm sure it is a low cost pool as it is set up for 1 algo and 1 coin.

Who is online

Users browsing this forum: No registered users and 1 guest