ethermine.org - The fastest way to mine ether

1356731

Comments

  • dr_pradr_pra Member Posts: 445 ✭✭✭
    Uncle rate varies a lot and 1000 blocks is just less than half a day of mining. As I already mentioned before you need to look on long term trends over several days to get a representative figure!
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    edited March 2016
    Yes I can: https://etherchain.org/statistics/miners

    At the bottom of the page you have an uncle statistics.I am not sure where you get your statistics and assumptions from.

    Also I advise you to follow the statistics for several days to get a feeling on how uncle rates vary over time before making any further claims.
  • zorvalthzorvalth Member Posts: 174
    edited March 2016
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    Then why does ethpool have a lower global uncle rate compared to dwarfpool despite having always a lower hashrate.

    In total dwarfpool has mined 232188 blocks and 15576 uncles which gives an uncle rate of 6.3%. Ethpool has mined 49502 blocks and 2369 uncles which gives an uncle rate of 4.6%.

    Better check your facts first before making any claims in regards to pool profitability.
  • workwork Member Posts: 2,084 ✭✭✭✭
    edited March 2016
    MrYukonC said:

    @bitcanuck cc: @dr_pra

    Bitcanuck,

    The thing you need to realize is that with ethpool at least, the 1% fee is essentially negated due to the way they pay out uncles. Very frequently I receive payouts greater than 5 ETH and @dlehenky can confirm this as well.

    So while you're "technically" paying 1%, you're actually being paid a bonus because of their uncle payment system. Make sense?

    It doesn't make sense statistically or mathematically tho... There is a reason the payout diff is ALWAYS higher then the network diff since uncles started getting paid on top of blocks. Sure, if you're lucky to hit one of the uncles, it works out (and I've hit more then "my share" of uncles on ethpool, so I'm no complaining), but, it doesn't make it more profitable.
    dr_pra said:

    Then why does ethpool have a lower global uncle rate compared to dwarfpool despite having always a lower hashrate.

    In total dwarfpool has mined 232188 blocks and 15576 uncles which gives an uncle rate of 6.3%. Ethpool has mined 49502 blocks and 2369 uncles which gives an uncle rate of 4.6%.

    Better check your facts first before making any claims in regards to pool profitability.

    Maybe this is because of the distribution of pools running broken clients that don't include other people's uncles properly? Having a low uncle rate is not necessarily a good thing right now - there is a decent chance it means uncles aren't getting paid out at all, and does not mean that more orphaned blocks aren't being found. This is well documented in geth issues. Having a higher uncle rate as a pool right now probably means more profits..... Maybe you should check your facts before telling other people to check their facts?

    Is ethpool running a beta client that is patched so that uncles aren't being missed? ...

    --

    That said, you're right that dwarf is keeping a lot of pieces for themselves... uncle inclusion rewards and fees aren't paid to miners.
    bitcanuck said:

    I get my statistical assumptions from Vitalik Buterin. Have you heard of him? :-)
    He calls it the last-block effect:
    "the miner that produced the last block “finds out” about the block immediately rather than after waiting ~1 second for it to propagate through the network, and thus gains an advantage in finding the next block"

    The assumption you are making based on that statement is honestly nonsense. Vitalik isn't exactly a genius either; he's had some absolutely terrible ideas during this project (just look at his original implementation of DAGGER... what a complete joke; or, further to that the "threat" of PoS to prop up the project's terrible PoW algo that stands to benefit more then scrypt from ASIC implementation, or even FPGAs with shared DRAM... there's plenty more examples too).
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    ethermine and ethpool both run the patch that fixes the broken uncle handling with geth. But in order to be 100% effective more than 50% of all miners need to upgrade.
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    edited March 2016
    BTW ethermine mines now more than 3 blocks per hour. This means the pool is now large enougth to guarantee fast and stable payments which should encourage more conservative miners to switch. Thank you all for your great support during the first days of the pool!!!
  • BiodomBiodom Member Posts: 693 ✭✭✭
    dr_pra said:

    ethermine and ethpool both run the patch that fixes the broken uncle handling with geth. But in order to be 100% effective more than 50% of all miners need to upgrade.

    @dr_pra Very much interested in what patch you are talking about? Would like to be one of needed >50%
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    just run the latest develop, it has the patch already included.

    see this issue for further details: https://github.com/ethereum/go-ethereum/issues/2298
  • zorvalthzorvalth Member Posts: 174
    edited March 2016
    Today the pool is insane :) With 0 luck I should mine around 0.55 Ethers per hour. Last night(european time) was a little negative as you can see the screenshot but today went insane! I love it! The hash is already 33-34GH(6 blocks/h) which is twice more then yesterday and as @dr_pra stated its enough for steady average payments for 24 hours regardless the luck.

    https://i1.someimage.com/66tlTYj.png
  • workwork Member Posts: 2,084 ✭✭✭✭
    dr_pra said:

    ethermine and ethpool both run the patch that fixes the broken uncle handling with geth. But in order to be 100% effective more than 50% of all miners need to upgrade.

    Right, so let's say you run the fix but dwarfpool doesn't. You'll.be including more uncles for dwarf then they include for you. Therefore dwarf's uncle rate will be higher. In that case, having a higher uncle count is a good thing!

    Essentially tho, uncle block percentages are very meaningless right now. And by running the develop branch other pools will benefit from more paid uncles, while your pool only benefits for some self-claimed uncles.
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    No, right now ethpool will essentially only include it's own uncles through that mechanism as other miners don't announce their uncles under certain conditions.
  • workwork Member Posts: 2,084 ✭✭✭✭
    @dr_pra interesting if true... that wasn't my reading of the problem, code, and patch.
  • blueboxbluebox Member Posts: 181 ✭✭
    @zorvalth So exactly where are you getting that summary showing E/h and span? Worker payout page only shows Tx after the Amount column... perhaps @dr_pra can enlighten us, I'd certainly like to see this in the summary. Really like the pool, haven't had a single hitch since switching over.
  • BiodomBiodom Member Posts: 693 ✭✭✭
    edited March 2016
    @work @dr_pra
    OT: Nothing to do with ethpool or ethermine, but is there something wrong with current geth/ethminer combo without this patch? qtminer works fine on the ethpool. Anybody have numbers that confirm or contradict this?

    I got one solo block 8days ago and 2 uncles 7 and 6 days ago respectively.
    Since then-nothing, although I should be getting a block every ~3.5 days on this 75-77mh rig.
    Waiting for Godot (almost)-just want to get one so I can switch to ethermine in good conscience.
  • zorvalthzorvalth Member Posts: 174
    edited March 2016
    bluebox said:

    @zorvalth So exactly where are you getting that summary showing E/h and span? Worker payout page only shows Tx after the Amount column... perhaps @dr_pra can enlighten us, I'd certainly like to see this in the summary. Really like the pool, haven't had a single hitch since switching over.

    I asked friend of mine to make a chrome extension which make some nice touches to the webpage but i got some criticism after i shared it because its not verified.

    The other important thing is the red sign for active workers and the averege minied ether per hour based on the last 24-30 hours

    https://i1.someimage.com/nddoSwE.png

    Here is the last version if somebody is brave enough to try it:)

    https://copy.com/domGMPwQqn0P
  • AraAra Member Posts: 8
    I'm really liking this pool, great job, just one quick question.

    Q) I'm running eth-proxy on one of my rigs. On both my rigs I connect to this with ethminer --farm-recheck 200 -G -F http://:8080/ however I don't see the actual name of the rig in my stats on the pool web page. I also only see one 'default' connected client. Does eth-proxy not send the name over to ethermine?

    I do see that I'm generating Ether, and I think the combined hashrate adds up, just wondering why it only shows one connected without a name?
  • AraAra Member Posts: 8
    Ara said:

    I'm really liking this pool, great job, just one quick question.

    Q) I'm running eth-proxy on one of my rigs. On both my rigs I connect to this with ethminer --farm-recheck 200 -G -F http://:8080/ however I don't see the actual name of the rig in my stats on the pool web page. I also only see one 'default' connected client. Does eth-proxy not send the name over to ethermine?

    I do see that I'm generating Ether, and I think the combined hashrate adds up, just wondering why it only shows one connected without a name?

    Nevermind, think I figured it out... there is an option in the proxy for namedworkers, missed that before, ethermine's download has it disabled by default.
  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    @Ara, It's better to run eth-proxy on each of your rigs and make them independent. That way if one goes down it's not a problem for your others that may be pointing at it.
  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    work said:

    MrYukonC said:

    @bitcanuck cc: @dr_pra

    So while you're "technically" paying 1%, you're actually being paid a bonus because of their uncle payment system. Make sense?

    It doesn't make sense statistically or mathematically tho... There is a reason the payout diff is ALWAYS higher then the network diff since uncles started getting paid on top of blocks. Sure, if you're lucky to hit one of the uncles, it works out (and I've hit more then "my share" of uncles on ethpool, so I'm no complaining), but, it doesn't make it more profitable.
    @work why not? Say for estimate sake there is a 10% uncle rate so you hit an uncle approximately every 10th round. You collect and extra >2.5Eth but have paid only 1% of 5eth *10 = 0.5eth in fees. So at a minimum you're 2eth up!
  • happytreefriendshappytreefriends Member Posts: 537 ✭✭✭
    I haven't hit an uncle in 14 rounds..
  • workwork Member Posts: 2,084 ✭✭✭✭
    @o0ragman0o the required credits to get a payout on ethpool are averaging 10% higher then network diff. Hitting an uncle probably averaging 1/20 blocks might balance that out, but really just adds randomness and variance to payouts.

    That said, I've had way better then expected luck getting paid for orphans on ethpool, so it'd worked out for me. Others, like @happytreefriends, aren't so lucky.
  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    @work, Still a lot of people like it and I think ethpool is a great option in the mix of things even if I don't really agree with it's statistics. The uncle payouts just add a little 'gameiness' fun without dropping certainty of payouts.
  • dr_pradr_pra Member Posts: 445 ✭✭✭
    Could you please move this discussion over to the ethpool thread, otherwise new users will be confused about the payment scheme if the read the thread. Thanks!
  • dlehenkydlehenky Member Posts: 2,249 ✭✭✭✭
    @work I've been very fortunate with uncles, having gotten 2 in a day 3 times, and 3 in one day once (2 where small, though). But dr_pra did say that the uncle policy does increase the average credits needed to get a block above the network difficulty. The uncle policy was put in place to help small miners who were waiting days to get a block, only to have it be an uncle. Overall, I think the policy is a positive for the pool.
  • AraAra Member Posts: 8

    @Ara, It's better to run eth-proxy on each of your rigs and make them independent. That way if one goes down it's not a problem for your others that may be pointing at it.

    Thanks, that's a good idea. Follow up questions.

    - Are most running the proxy or using qtminer?
    - When I use the options `ethminer --cl-local-work 256 --farm-recheck 200 --cl-global-work 8192 -G -F http://IP_ADDRESS:8080/RIGNAME` on my non-main rig running with a GeForce GTX 760(OpenCL 1.2 ) I see a lot of 0 H/s but every 2-3 logs I'll see around 10034220 H/s. Is this normal? I don't see this on my r9 380 main rig. What options are people using here for both a r9 390 and Nvidia cards?

    miner 06:29:47.899| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.209 s miner 06:29:48.113| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.213 s miner 06:29:48.325| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.211 s miner 06:29:48.534| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.209 s miner 06:29:48.744| Mining on PoWhash #f9239192… : 10034220 H/s = 2097152 hashes / 0.209 s miner 06:29:48.952| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.208 s miner 06:29:49.164| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.211 s miner 06:29:49.374| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.21 s miner 06:29:49.589| Mining on PoWhash #f9239192… : 9799775 H/s = 2097152 hashes / 0.214 s miner 06:29:49.799| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.209 s miner 06:29:50.009| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.21 s miner 06:29:50.223| Mining on PoWhash #f9239192… : 0 H/s = 0 hashes / 0.214 s miner 06:29:50.434| Mining on PoWhash #f9239192… : 9939109 H/s = 2097152 hashes / 0.211 s
  • o0ragman0oo0ragman0o Member, Moderator Posts: 1,291 mod
    edited March 2016
    @ara, I'm not really one to be able to answer that (maybe @Genoil could) but just to test, set your farm-recheck to 2000. It will then have a far larger sample of hashing to average over for it's MH/s calculation. If that number makes sense, then your probably ok.

    BTW, are you using Genoil's CUDAminer?
  • AraAra Member Posts: 8
    @o0ragman0o My GTX Rig is a hackintosh, all the cudaMiner stuff I could find is compiled binaries for Windows. Anyone have a link to the CUDAMiner source so I can build for OSX? @Genoil ?

    Spot on with the farm-recheck, I'll play with the numbers a bit but gonna wait to see if I can get CUDAMiner running on OSX first.

    miner 07:31:33.243| Mining on PoWhash #9d135cd8… : 4161015 H/s = 4194304 hashes / 1.008 s miner 07:31:34.250| Mining on PoWhash #9d135cd8… : 2084644 H/s = 2097152 hashes / 1.006 s miner 07:31:35.258| Mining on PoWhash #9d135cd8… : 2080507 H/s = 2097152 hashes / 1.008 s miner 07:31:36.263| Mining on PoWhash #9d135cd8… : 2086718 H/s = 2097152 hashes / 1.005 s miner 07:31:37.273| Mining on PoWhash #9d135cd8… : 2078445 H/s = 2097152 hashes / 1.009 s miner 07:31:38.278| Mining on PoWhash #9d135cd8… : 4173436 H/s = 4194304 hashes / 1.005 s miner 07:31:39.283| Mining on PoWhash #9d135cd8… : 2088796 H/s = 2097152 hashes / 1.004 s miner 07:31:40.289| Mining on PoWhash #9d135cd8… : 2086718 H/s = 2097152 hashes / 1.005 s
  • GenoilGenoil 0xeb9310b185455f863f526dab3d245809f6854b4dMember Posts: 769 ✭✭✭
    @Ara eth mining on Windows+Nvidia is best done on win7/8 or by downgrading drivers. Your hashrate could be around 12MH/s. AFAIK nobody has successfully built my fork on OSX.

    Src here:

    https://github.com/Genoil/cpp-ethereum
  • ethfanethfan Member Posts: 458 ✭✭✭
    edited March 2016
    @dr_pra, your site says 5 invalid shares is cause for concern. Is it not 5% of submitted shares? What issues need to be looked at for high invalids? Mainly latency problems? Aggressive overclocking/undervolting? Sub-optimal parameters when launching miner?
Sign In or Register to comment.