Bitcoin Forum
April 24, 2024, 02:36:14 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Is single algo hub mining still efficient, with lots of multiminer helpers now?  (Read 128 times)
cosmicog (OP)
Newbie
*
Offline Offline

Activity: 6
Merit: 0


View Profile WWW
February 20, 2018, 06:05:08 PM
Last edit: February 20, 2018, 06:25:03 PM by cosmicog
 #1

I'm trying to write multi algorithm pool switcher code for miningpoolhub on Linux.
It's gonna be Open Source  Wink But I have no experience about pool switching efficiency.
And I have some questions.

I was about to try both of options, but I need to do test for a long time to see their profits.
which is too hard. So, which one of below options is more efficient way to mine?

Option 1. Starting miners on specific pool ports which is assigned to that coin,
when I see that coin on top of profitability list. For example, if it's Zcash, My program will run
Zcash on bminer, If it's Zcoin, My node will run Zcoin on ccminer.

Option 2. I won't care about coin, I'll just check the top algorithms on the list.
If it's Equihash, I'll run bminer with Equihash hub worker, If it's Skein, I'll run ccminer on hub's skein port.

My miners are sometimes struggling about stratum difficulty. And before the miner finds best stratum
and initializes itself to get good hashrate, BAMM best coin or algo is changing to other one, and I'm seeing
that If I would continue mining previous coin, I would get more profit. That's why I'm scaring from hub feature.
Because it makes ALL AUTO SWITCHER FELLOWS to go there and makes crazy difficulty changes in a minute.

I've seen even with normalized profit column, it happens. I think it's happening because, that coin or algo
needs time for normalization. And then We see them on top of list, and after a bit time for normalized profit,
it's being less profitable again! Also I think because of other auto switcher mining helpers Cheesy But, does it still
profitable, mining for 2 minutes? Or should I don't look at normalized profit?

As a SUMMARY;

Option 1.1 - Mining specific coin on it's own pool by not normalized profit data.
Option 1.2 - Normalized one of Option 1.1

Option 2.1 - Mining Specific algo on algo hub by not normalized data
Option 2.2 - Normalized one of Option 2.1

Option 3 - EXPERIMENTAL IDEA, I'll say, Do not go to the best, for example go to the second coin's
own pool first. Then if norm. profit of best is stable for a while, go there in 5 mins. Also I'm thinking
about showing favor to specific algo defined by user in config file, for example, if the value doesn't
deserve switching that much, stay at coin or prefer that coin.

I didn't write "Option 3 - Mine Zclassic 4ever" as a joke.
Because I don't want some redditer to come here and say:
"Everything changes, you can't decide on bla bla market bla bla
Buy ASIC Bla bla" or something.
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!