Креирање на mining pool

Дискусија во форумот 'Дигитални Крипто-валути' започната од blackelf, 21 ноември 2017.

  1. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Немав што да правам и добив идеја да направам mining pool. Да се обидам да направам де :).
    Е сега ме интересира делот кај VPS. Каква машина да барам? Со подобар процесор или рам?

    Исто така ме интересира ако успеам да дигнам pool за Electroneum, дали некој од вас би префрлил некоја машина?

    Ме интересираат ваши мислења и совети. Слободно пишете што мислите.
    И секако, секоја помош би била добредојдена.
     
    На stefan_off и crazzyy84 им се допаѓа ова.
  2. HocusPocus

    HocusPocus Gaining Experience

    804
    399
    3 Јули 2007
    Одржувањето е pain in the ass. Ќе се најде некоја будала ќе те сруши со ботнет или DDoS.
     
  3. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Знам дека е главоболка, ама јебига.. немам што попаметно да правам.

    Sent from my Le2 using Tapatalk
     
  4. Alexander_mkd

    Alexander_mkd Intern

    15
    2
    20 Април 2012
    Машко
    Ram16 gb minimum ti treba a i procesor treba soliden.Problemot ke ti e odrzuvanje. Treba da si dobar u shell, zoso mn bagovi se javuvaat a i ko si ti rekoa ima sekakvi budali :)

    Испратено од HTC Desire 626 преку IT.mk форум апликација
     
  5. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Читав до сега и се едуцирав (sort of)
    upload_2017-11-21_21-49-40.png
    Јас би пробал со 2GB рам и би идел со Vultr серверите.
    Некој од искуство да ми каже подобар хост со многу многу добра DDoS протекција?
     
  6. HocusPocus

    HocusPocus Gaining Experience

    804
    399
    3 Јули 2007
    Од искуство колку повеќе конекции толку повеќе рам меморија троши и треба еден куп твикувања на самиот систем како open files limits, connection limits итн. Можеш да пробај солиден ВПС од линкот во потпис за 9 евра. Исто и диск ти треба минимум колку што е големината на блокчеинот за да инсталираш демоните ...
     
  7. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Го проверив тој од твојот опис. Твој е? Со заштити како стои?

    Sent from my Le2 using Tapatalk
     
  8. HocusPocus

    HocusPocus Gaining Experience

    804
    399
    3 Јули 2007
    За тие пари не наоѓаш ништо приближно. 12 гига рам и 300 гига ССД. И до сега не сум приметил ддос на серверите ниту пак некогаш да е бил паднат. Ги користам 5-6 месеци.
     
  9. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Ама сепак, нели е полесно пул да се бутне отколку страна?

    Sent from my Le2 using Tapatalk
     
  10. HocusPocus

    HocusPocus Gaining Experience

    804
    399
    3 Јули 2007
    Сеедно дали е пул или страна, има многу филозофии ама додека не почнеш да си играш нема да научиш.

    Пробај со nodejs-pool да го прилагодиш за електрониум.
     
  11. Whisky Vishera

    Whisky Vishera Practice makes perfect

    1,145
    1,806
    18 Март 2015
    Ова е за монеро, ама важи за која и да било друга криптовалута.

    You freshly discovered this coin with all the hot news with darknet.

    You started to connect on poloniex and saw that in few weeks XMR price multiplied 3 times.

    Even now bigger price than the so famous Ethereum.

    And all that with real anonymity and fungibility.

    So you decided to run your own mining pool.

    You want to be part of the gold rush.
    And smarter than the mass you want to provide pool as formely smarter one were selling showel to dig the ground.

    Fair Enough.

    Bad news: you are two years late.

    Anyway you have an IRON will power.

    If your plan is to rent a $5/month Ocean digital linux instance for that
    while buying a $100 uber cool .io domain name, you MUST QUIT NOW. REALLY

    So if you are ready to become big, you have to think big.
    If you want to become rich you have to be rich. erm...
    Or at least ready to spend more money on this project.

    And associate with a very good PR.
    PR is the key point with trust and good infrastructures.

    --Servers and network:

    At begining everything is easy.
    Few miners, few shares to validate,
    Lousy connexion with packet lost,
    you feel so proud to have mined your first block and send your first payement.

    Fine fine.

    It's been 3 weeks you were looking at your log with high anxiety,
    expecting this fucking message : BLOCK FOUND.

    Some pools admins don't even dare to look at those log, they know they find every three blocks in the blockchain.
    They have hash rate, huge power computation from thousand of miners, and more thousand of botnet.

    Their concern is not finding a block, but is more about how to keep finding a block!

    That the mindset shift.

    The first pitfall: ulimit.
    The name of this is quite ironical, as far as you will reach the standard limit of opened file by a user.
    The main effect is that the pool cannot accept new connexion, miners will be unable to work on your pool.
    They will move away after discovering they loose few hours of precious hash on your lousy setuped pool.

    And be sure of something, they will never come again ! [​IMG]

    So Second advice: read carefully the manual of ulimit, and try to understand how you can increase the default limit.

    Alright, you avoid the ulimit issue, or solved it fast enough to not loose all your hashrate.
    Seems trust is growing and with you PR doing good job on bitcointalk to attract all those new
    miners thinking they will become rich recycling their GPU RIG catching dust in their basement
    after mining a few microbtc few years ago.
    They find a good excuse to switch on the power, and make their wife agree for maintening >40c inside the house
    "WE ARE GOING TO BE RICH MUMMY, MONERO XMR IS THE NEW BITCOIN"

    Well, why not. They will not be the first to give a try...

    So seems your pool start to grow hashrate.

    Good

    You start to become a target.

    [​IMG]

    yes

    From competitors... you know those other pools owner lurking on your hashrate.

    What's the best way to steal your hashrate ?

    Well, easy too : DDOS

    When you get DDOSed, your miners cannot connect to your pool to get work.
    They switch to another pool.
    With little luck, they switch to the pool which owner paid for ddosing you [​IMG]

    See the point [​IMG]

    So what is solution [​IMG]?
    You might think, for few bucks your VM provider propose cloudflare solution.......
    Lucky you.

    You have to change of provider.

    Cloudflare deals very baldy with stratum and stratum like protocol.
    Often blocking it, as malformed http traffic.

    YOU ARE STUCK.

    The BEST and ONLY solution is Arbor Network.
    With Arbor you can deal with 300GB/s DDOS without one timeout of your connected miner.

    So choose wisely your hosting solution, and check if they propose Arbor anti DDOS solution.
    This can cost quite big depending of their plan.

    But running a famous and big pool has no Price !!!!

    So now, you can deal with ulimit and DDOS.
    GREAT DUDE.
    I hope you didnt let your reddis instance listen to the world. !!!!! [​IMG]
    Someone would be kind enought to change some value on currentvalidsharecountbyaddress
    and steal all your coins !

    [​IMG]

    Fuck your payout thread start to fuck around, and you didnt launch it separated from the rest of code.
    Damn you should have read better the fucking pool doc, and start to write a screen/tmux launcher with
    separating all the thread from each other.
    ANY MINUTES OF DOWNTIME IS DAYS OF PR FOR GETTING BACK MINERS TO YOUR POOL.

    As thousand of share to validate flow to your pool per second.
    You start to see your CPUs working more and more.
    Often in the RED near to 99%.
    Your server is to slow to validate share.

    Or not.
    May be the share validation code is bloated and use too much cpu for this little task....
    It's time to look at your pool wallet balance. And search for a developper known for
    Optimizing code.
    For quite a bunch of XMR may be he will be able to write you a piece of code in C/asm,
    which will validate your share in 100 times less time than standard pool code....

    who knowzzz [​IMG]?

    Damn, was it worth the spend ? now your cpu are in green zone and your share validation rate limit
    is near the moon.

    Btw, some smarted ddoser, can start to overflow you not on network level, but on cpu usage.
    Submitting too much bad share, consumming a lot of CPU and precious time on your server.
    So much that limit are reached and your miner's share are not validated anymore.

    May be missing some valid block in the process.

    That's wild world !!! you have to really keep that in mind all the time.

    So now, you start to have some miners, mine few block a day, but some of your block are orphaned.

    What does it means orphaned ??

    Well while your block was valid, the network rejected it and choose another block for the same place
    in the blockchain.

    Lot of pools can find a block candidate for the same height.

    So not only you have to find a block(well not you, but your miner).
    But to be sure to be rewarded, the network has to accept it, faster than concurrent blocks submitted by competitors.

    What [​IMG] you didn't know [​IMG] But, this is how decentralised consensus works my dear !!!!

    Well in fact it's even a little bit more complicated.
    The blockchain has often several sub branchs of blocks to choose starting from the latest validated and accepted block.

    So the network consensus algorithm will choose the sub branch of blocks whith the heaviest added difficulty.

    That's the core protocol algorithm.

    You can find a block, start to work on the next one, and one hour later you have a blockchain reorganisation.
    Because another branch was choosen by the network, and therefore you loose all the MONERO you get rewarded
    from your newly orphaned block ! [​IMG])) (This is a wild world, remind it).

    So what strategies can you use to increase your block acceptance [​IMG]?

    An easy one is to increase the number of peers which your mining bitmonerod node is connected too.

    This is the starting of the the fun.

    You have to patch code and compile again. ! [​IMG]

    Default number is 8, which means your bitmonerod will be connected to 8 other peers at maximum.

    Why not you increase it to 100 ? or 1000 ? you don't fucking care you overcame the ulimit pitfall.

    Second, why not grow your number of nodes ?
    The more node you run, the more weight you have on the consensus algorithm.

    What if you run 10, 20 or even 50 bitmonerod instances on all continents to covers all the space.

    With 20 nodes spreaded around the world, directly connected to your primary node and therefore
    relaying your fresh block in the network, you increase the chance for it to be fully accepted.

    Today, given the little number of pools, there is not so much competion in spreading your block.
    So having a few secondary node as prefered peer from your primary node, all of them with increased max peer number
    can give good result in consensus acceptance, avoiding orphaned block.

    It can also increase your luck !!!

    What is luck [​IMG] well vast question.
    But in mining, with a given difficulty, you can estimate from globalhashrate and you localhashrate, the time to mine the next block.
    Luck is positive if takes your pool less time for your pool to find the next block, than expected.
    It is negative if it takes more time than expected.

    Remind, mining is a lottery, the time finding the good number can be statistically estimated, but it is not reality.

    So if you have fast connexion to supply and receive work from miners,
    fast share validation reducing the processing time on pool
    you will discoved faster than expected if a share leads to a valid block.
    Having a bunch of node to spread it through the network, will increase acceptance factor in blockchain.
    Therefore your luck increase !
    Your miner are happier!
    Your PR has easier job to attract more miners!
    Your pool wallet has grown far bigger than you could imagine!
    Now you can play the big whale in poloniex troll box !!!!

    To conclude:
    NEVER FORGET That reputation is very HARD to build.
    And very easy to destroy.

    This is not your fucking money, but other's one.
    Be ready to have to put money from your pocket to pay your miner in case of any trouble.
    Some pools miner loose their profit of 3-4 months just to compensate miners after big trouble with pool's wallet.

    Miner are easely talkative (irc/bct), and share their infos.
    Act bad one time and you loose everything.

    And you are still thinking about a VM in amazon instead of digital ocean. GET FUCKING LOST!!!!!
     
    prichina, blackelf и на M.Marlon им се допаѓа ова.
  12. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Пекол кажано. Кога викам пекол мислам на јако, прејако. Идеата ми е да вложам малку на почеток и да видам како ќе оди првиот месец. 1% fee за пулот и опционално уште 1% за донации или %0.5 со %0.5. И да има некој реферал програм. Пример донесеш одреден број на хеш со код и добиваш помал fee, па и некој бонус. Уште се размислувам дали да почнам. Гледам на лоши и добри страни, ама реално што би изгубил?

    Sent from my Le2 using Tapatalk
     
  13. crazzyy84

    crazzyy84 Gaining Experience

    235
    153
    26 Април 2017
    Машко
    нерви

    Испратено од Lenovo K33a48 преку IT.mk форум апликација
     
    На DrMTR и EmTe им се допаѓа ова.
  14. HocusPocus

    HocusPocus Gaining Experience

    804
    399
    3 Јули 2007
    Најдобро е да почнеш пул на нов коин, невозможна мисија е да почнеш пул без трафик за популарен коин.
     
    На EmTe му/ѝ се допаѓа ова.
  15. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Ги имам. Истите ко и за Курир ;)

    Сметам дека ETN е уште релативно нов коин и уште многу машини и хеш ќе се префрли од етериум и други валути на Electroneum.
     
    На crazzyy84 му/ѝ се допаѓа ова.
  16. crazzyy84

    crazzyy84 Gaining Experience

    235
    153
    26 Април 2017
    Машко
    дури денес ги зедов парите замисли која трилогија

    Испратено од Lenovo K33a48 преку IT.mk форум апликација
     
  17. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Не е некој проблем де. 21 век, парите за две недели. А замисли пред еден век како работел системот? Па да, прилично исто..
     
    На crazzyy84 му/ѝ се допаѓа ова.
  18. blackelf

    blackelf Gaining Experience

    359
    248
    14 Декември 2013
    Машко
    Сепак се одлучив за ovh.ie.
    За Anti-DDoS вика: 3 Tbps hosted in 9 OVH data centres.

    Некој искуства со OVH да сподели?
     
  19. Psycore

    Psycore Gaining Experience

    193
    119
    12 ноември 2013
    Мислам дека не е толку лесно да го изведеш ова, колку што мислиш. Лично немам искуство со вакво нешто, но по муабетов можам да кажам дека треба познавање во неколку области и технологии. Немаш наведено колку си спремен да инвестираш, но не е само базичен VPS пакет.

    И да успееш све, на крај следува маркетинг делот, за кој исто така треба да фрлиш пара. Како ќе ги привлечеш сите тие луѓе да копаат баш кај тебе?

    Јас на твое место би ги инвестирал тие пари long-term ве некој коин. Но, како што кажа дека си загреан за ова и имаш слободно време, пробај си ја среќата. Барем ќе научиш неколку IТ и бизнис лекции за во животот. :)
     
  20. Whisky Vishera

    Whisky Vishera Practice makes perfect

    1,145
    1,806
    18 Март 2015
    @blackelf ме интересира направи нешто во врска со ова?
     
    На blackelf му/ѝ се допаѓа ова.
  21. HocusPocus

    HocusPocus Gaining Experience

    804
    399
    3 Јули 2007
    На blackelf му/ѝ се допаѓа ова.

Сподели