[ANN] [QRK] Quark | Core 0.10 upgrade

[K1773R]

Thanks, I got it to work. I didn't see the README. How do you know when you've found a block?

Code:

accepted: 1/1 (100.00%), 896.91 khash/s (yay!!!)

[auctioneeeeer]

Thanks, I got it to work. I didn't see the README. How do you know when you've found a block?

Code:

accepted: 1/1 (100.00%), 896.91 khash/s (yay!!!)

I have compiled in Linux and used the precompiled windows versions here, and they are abysmally slow for me.  On a single core, the wallet gives me 120Kh/s.  The cpuminer version gives me about 6Kh/s.  I've tried the AVX and AVX2 versions on both Ivy-Bridge and Haswell Xeons to no avail.  Using no options on the command line, and --benchmark gives me the same slow h/s.  Any ideas?

[Lauda]

Price too low 

[Lauda]

Price too low 

Meh, it's better than .000001 or .0000015 which it was at earlier.

Meh, meh, you're somewhat right.

[auctioneeeeer]

Thanks, I got it to work. I didn't see the README. How do you know when you've found a block?

Code:

accepted: 1/1 (100.00%), 896.91 khash/s (yay!!!)

I have compiled in Linux and used the precompiled windows versions here, and they are abysmally slow for me.  On a single core, the wallet gives me 120Kh/s.  The cpuminer version gives me about 6Kh/s.  I've tried the AVX and AVX2 versions on both Ivy-Bridge and Haswell Xeons to no avail.  Using no options on the command line, and --benchmark gives me the same slow h/s.  Any ideas?

Make sure you have -a quark added to the line that runs minerd.

I knew I was missing something simple, what's your qrk address, let me hook you up with some coins!

[Neisklar]

For those wanting to know where their processor will be in terms of how fast, here's some stats:

...snip...

I think higher GHz, more khash you'll get.
Also, on Linux:

cat /proc/cpuinfo

Even more important: 64bit miner!
Roughly double speed.
My older Core2 Laptop Windows doubled its speed, as well as an Xeon CPU running Ubuntu

@neisklar, is the p2pool source @ github? want to run one localy

Not yet, that needs definatly some workover as i have hundreds of ourcommented debugging lines in it. That hardwiring of the DONATION_SCRIPT into the code and !!! into the validation of the p2pool's internally blockchain !!! was hard to get. But is/was a clever thing. If you wan't to join an existing p2pool network you can't change that DONATION_SCRIPT, since then for that p2pool note the blockchain would be invalid...

[Neisklar]

So what does this mean when -H is added:

Hash:   0000005ecbfa21bbe690a9467424cfe6282e5387ab6ff50b5e9d89a89c1f7eed
Target: 0000000181fe0000000000000000000000000000000000000000000000000000
[2013-08-17 20:42:19] DEBUG: hash > target (false positive)
Hash:   000000aa5d6b11e1b603c5add952188f2b50c566f3cc0aeafcbc9dc89f3d1a26
Target: 0000000181fe0000000000000000000000000000000000000000000000000000
[2013-08-17 20:42:38] DEBUG: hash > target (false positive)
Hash:   000000957e7b92f87ae5e17db8732fdcfe921afbbac70e28bfd8765c1913c494
Target: 0000000181fe0000000000000000000000000000000000000000000000000000

??

It's some debug output, i used to see during the tests if anything goes the right way.

You can even omit the -H since even without you should see speed messages.
IF you want it quit, and only print out accepted/rejected and global speed usw -Q

Also, why doesn't my address show up in the pool? http://176.221.46.81:8372/static/

I have 4/4 shared accepted.  

Please read about the workings of p2pool and especially the fake difficulty.
The links are somewhat up in one of my posts, or look in the bitcoin wiki

Edit: https://bitcointalk.org/index.php?topic=260031.msg2948763#msg2948763

[Lauda]

Just found a block with the new miner.

Oh nice, gratz!

[victzhang]

Just wanted to report back. Left a VPS on for 10 hours mining and got 6 blocks, I mined with the CPU miner for 5 hours and only got 1 block. Interesting thing is that I got some odd debug outputs with -H added but got a higher hash-rate.

Please remember that now the difficulty (200+) is much higher than before (~100)...

[victzhang]

No, this is not a new qt wallet, only a standalone miner that mines more efficiently than the miner built into the qt wallet. You keep your current qt wallet application and wallet.dat file the same.

See this thread for using the miner on the p2pool:

http://forum.quarkcoin.org/Thread-Quark-Mining-Pool-Creation-Needed?pid=74#pid74

It's basically:

Code:

minerd_xyz.exe [b]-a quark[/b] -t <numthreads> -o http://176.221.46.81:8372 [b]-u <youquarkaddress>[/b]

To get some more details about your speed add an -H

This pool takes cucrrently 3% fee, which you can verify yourself:
http://176.221.46.81:8372/fee

Please also inform you about how p2pool works, esp about that fake difficulty:
https://en.bitcoin.it/wiki/P2Pool#Payout_logic
https://en.bitcoin.it/wiki/P2Pool#Frequently_Asked_Questions

----snip---

Putting the hash_init-stuff outside of the loop isn't gaining anthing in cpuminer

I am a P2Pool newbie and I have a problem using this pool. I have a machine running at 1000khash/s, but the payout is very unstable. I am getting the "accepted" messages almost every minute. However, sometimes I get zero payout for a few dozen minutes, and my address is even not listed on the pool website during this period

I have this issue on both Windows (use the precompiled w64 binary) and Archlinux machines. Is it normal? Perhaps I don't understand the P2Pool payout logic correctly???

Any help will be greatly appreciated!

[Neisklar]

Please read the entire p2pool entry in the wiki linked above.

But in short:

We have 3 difficulties/targets:

1) Target for Block: The Network difficulty, the highest of all three.
2) Target for a share: mostly the same as in other pools: The amount of shares calculates you reward. (Lower than network target, retargets so that the entire p2pool generates a share around every 15 seconds.
3) Target for your miner: That is even lower, So the pool eseential 'lies' to your miner. That has a practical reason, the pool has a better way to get you hashrate. This means the 'shares' you submit aren't all the shares that count as reals shares.

Edit: current share diff is: 0.113370, and fake diff is 0.008985

For reward calculation the last n shares are used: n= 5750 pool shares or the average work to solve 30 blocks (whichever is smaller).

In the statistics page all solved blocks are linked to the blockexplorer, so you can look who gets how much.

BTW:there are some ways to change the fake and share diff buy adding something values on the username, i can't currently find that info, so i will edit in in later.

Edit:
you can add a value with '+' at the end of your username, to set the fake diff to that value.
you can add a value with '/' at the end of your username, to set the share diff.

That / setting is good for highpower miners, since that keeps the share-diff in the network low.

-------------------------------

Who is QjNp65rfy4FENdYNN8Z66pP5NCCiWbpa5K
Please PM me

[K1773R]

@Neisklar: still waiting for a true p2pool src are you working on it or is this no goal?

[victzhang]

Please read the entire p2pool entry in the wiki linked above.

But in short:

We have 3 difficulties/targets:

1) Target for Block: The Network difficulty, the highest of all three.
2) Target for a share: mostly the same as in other pools: The amount of shares calculates you reward. (Lower than network target, retargets so that the entire p2pool generates a share around every 15 seconds.
3) Target for your miner: That is even lower, So the pool eseential 'lies' to your miner. That has a practical reason, the pool has a better way to get you hashrate. This means the 'shares' you submit aren't all the shares that count as reals shares.

Edit: current share diff is: 0.113370, and fake diff is 0.008985

For reward calculation the last n shares are used: n= 5750 pool shares or the average work to solve 30 blocks (whichever is smaller).

In the statistics page all solved blocks are linked to the blockexplorer, so you can look who gets how much.

BTW:there are some ways to change the fake and share diff buy adding something values on the username, i can't currently find that info, so i will edit in in later.

Edit:
you can add a value with '+' at the end of your username, to set the fake diff to that value.
you can add a value with '/' at the end of your username, to set the share diff.

That / setting is good for highpower miners, since that keeps the share-diff in the network low.

-------------------------------

Who is QjNp65rfy4FENdYNN8Z66pP5NCCiWbpa5K
Please PM me

Thank you for your clear explanation!
Could you please tell me what share diff is good for a 1000khash/s miner if I add a value with '/' after my username?

[Neisklar]

@Neisklar: still waiting for a true p2pool src are you working on it or is this no goal?

Have cloned the source git, but not worked on it. But will come.

Thank you for your clear explanation!
Could you please tell me what share diff is good for a 1000khash/s miner if I add a value with '/' after my username?

Honestly, i have no idea:)

But lets try:
If my math is correct you need at 1000kHash average 16,77 seconds to get a share at diff 1 (2^24 hashes needed)
So to not push the diff, maybe you could use /2 or /4 (average 30/60 seconds), since the entire pools target is to have one share every 15 seconds absolute

The pool should of course credit that higherdiff shares more than lowerdiff shares.

[bronan]

On P2pool you see often blocks being found but that does not mean its a valid block, it can take a while before its accepted or orphan/not valid
Especially the high timeouts can produce lots of rejected
Thats why i have quit using P2Pool along time ago, all together lost several hundred of btc because of this failures ofcourse it has been improved a bit but i sure as hell never use it again.
 

[smolen]

Just wanted to report back. Left a VPS on for 10 hours mining and got 6 blocks, I mined with the CPU miner for 5 hours and only got 1 block. Interesting thing is that I got some odd debug outputs with -H added but got a higher hash-rate.

Please remember that now the difficulty (200+) is much higher than before (~100)...

Right, however, some of them were when it was high.

Time for coinchoose.com to get a page for CPU-only coins. Would be nice to compare profits generated by Quark, Primecoin and Memorycoin.

[bronan]

I forgot to mention that i do not think that the minerd has less mining result
With this high difficulty over 200 its has not given me any coins for 26 hours on both my main machines at the old gen from the client
I have a i7-3770k with running at its 4 cores and the amd fx8320 with 4 cores however i think more and more cpu miners are on the quark
With minerd 2.3.2 i see the cpu get heated a bit more
This far i see 71 to 76 kh/s on the i7 amd 59 to 62 on the fx8320.
For prevent excessive heat i let both machines run this way, because we have high temps over here

 

[Lauda]

Another day of mining QRK 

[Neisklar]

On P2pool you see often blocks being found but that does not mean its a valid block, it can take a while before its accepted or orphan/not valid
Especially the high timeouts can produce lots of rejected
Thats why i have quit using P2Pool along time ago, all together lost several hundred of btc because of this failures ofcourse it has been improved a bit but i sure as hell never use it again.
 

Did that happen on a p2pool network, or on a p2pool running as standalone pool?

Edit:
Also something weird is happening, since a certain point today, in the stats page, all shown block hashes are just the wrong ones.
Looking in the block explorer under the block number i see different shares, and also corrext payouts.

Maybe i need to reset the database.
If i will do that, than i will announce it way before. It shouldn't affect earnings.
Or maybe i just leave it, since it works, it's just the stats page that shows the wrong hash, the block number itself and the payouts are corrext

[K1773R]

On P2pool you see often blocks being found but that does not mean its a valid block, it can take a while before its accepted or orphan/not valid
Especially the high timeouts can produce lots of rejected
Thats why i have quit using P2Pool along time ago, all together lost several hundred of btc because of this failures ofcourse it has been improved a bit but i sure as hell never use it again.
 

Did that happen on a p2pool network, or on a p2pool running as standalone pool?

Edit:
Also something weird is happening, since a certain point today, in the stats page, all shown block hashes are just the wrong ones.
Looking in the block explorer under the block number i see different shares, and also corrext payouts.

Maybe i need to reset the database.
If i will do that, than i will announce it way before. It shouldn't affect earnings.
Or maybe i just leave it, since it works, it's just the stats page that shows the wrong hash, the block number itself and the payouts are corrext

no, he just dosnt understand the basics. what he sees as block is simply a p2pool sharechain share/block.