P2pool feathercoin linux

Wellenreiter

The number of workers on a node is not that interesting, as the p2pool cloud is one virtual pool with many nodes. Most p2pool nodes have a small number of miners only.

You can find a list of public p2pool nodes on http://p2pool.neoscrypt.de. It is dynamically generated.

With the -n parameter you specify one or more p2pool nodes to connect to initially after starting your node. This helps to connect your node to the p2pool cloud.
If you omit this parameter, your node uses a node specified in the code to create the initial connection. As soon as an initial connection is made it gets the addresses of more peers from the other node.

gwbraa

I read the website P2pool better now that I am not a Zombie.

Pretty nice idea it’s definitely what was looking for. It’s like mining in your own POOL.
I will try to get 100MH/s to keep for like one year in this node so more people maybe be interesting in mining using the P2pool

really nice work from you guys

Wellenreiter

@gwbraa : how man riggs do you have?

currently you provde 210 Mhash.

There is a little problem with that :(

The ‘normal miner’ has a hashrate of <= 200 khash, what is 0.1 % of your hashrate.
The result is, that your rigg currently finds blocks faster than the other nodes can provide valid hashes to their p2pool node and therefore never get paid for their mining efforts.

You can see that in the p2pool display.
Whenever the time to block is shorter than the time to share, miners with smaller hash rates have problems to submit shares before a block is found.
Whenever a block is found new work is distributed to the miners and their work is useless.

can you try the following:

configure your miner or miners to use a higher share difficulty than the default.
- you can do that by adding a +<number> to your miner’s user name.
- example: ccminer -N 60 -i 19.5 -a neoscrypt -o stratum+tcp://146.185.178.40:19327 -u YOUR_FTC_PAYOUT_ADDRESS+5 -p x

There is no space between the payout address and the ‘+’ and the ‘+’ and the number.

You can vary the number.

I hope, that will enable the other miners to submit shares before new blocks are found.

gwbraa

Diffitcult from the miners are 3555.52757900 right now

That’s help??? or not really?

I had over 200 miners connect

gwbraa

@Wellenreiter said:

@gwbraa : how man riggs do you have?

currently you provde 210 Mhash.

There is a little problem with that :(

The ‘normal miner’ has a hashrate of <= 200 khash, what is 0.1 % of your hashrate.
The result is, that your rigg currently finds blocks faster than the other nodes can provide valid hashes to their p2pool node and therefore never get paid for their mining efforts.

You can see that in the p2pool display.
Whenever the time to block is shorter than the time to share, miners with smaller hash rates have problems to submit shares before a block is found.
Whenever a block is found new work is distributed to the miners and their work is useless.

can you try the following:

configure your miner or miners to use a higher share difficulty than the default.

you can do that by adding a +<number> to your miner’s user name.

example: ccminer -N 60 -i 19.5 -a neoscrypt -o stratum+tcp://146.185.178.40:19327 -u YOUR_FTC_PAYOUT_ADDRESS+5 -p x

There is no space between the payout address and the ‘+’ and the ‘+’ and the number.

You can vary the number.

I hope, that will enable the other miners to submit shares before new blocks are found.

What’s a good speed to make the most of it?

gwbraa

I was founding a lot of blocks. but I do understand what you said. when I turn on my computer the miner he was not getting accept shares only a fews I guess because it’s too difficulty
I will keep 10 MH/s to see what’ will happen in 24 hours.

But when I was fast I was found a lot of blocks but I guess I was supposed to find even more no?

Wellenreiter

I have no idea.
you could start with +100 and see what will happen.
You should wait for 1 -2 hours to see a good result.
Then you can adjust the difficulty.
There are two different settings possible:

user name+>number -> gives a share difficult of number+ normal share difficulty
username/number -> gives a fixes share diff on number

I try to find the documentation so I can give a definite answer.

gwbraa

@Wellenreiter said:

I have no idea.
you could start with +100 and see what will happen.
You should wait for 1 -2 hours to see a good result.
Then you can adjust the difficulty.
There are two different settings possible:

user name+>number -> gives a share difficult of number+ normal share difficulty

username/number -> gives a fixes share diff on number

I try to find the documentation so I can give a definite answer.

The weird think it’s about the shares I cannot have one share. I reboot the server and still the same.

I will keep average of 10MH/s for some hours to see what happen

Wellenreiter

@gwbraa I have implemented ‘adaptive share difficulty’ on my p2pool server and just testing it.

It basically implements a different rule set to calculate the share difficulty on a per worker basis.

if you want to try : https://github.com/wellenreiter01/p2pool-neoscrypt
use the ‘adaptive’ branch, not the master branch.

In order to enable the adaptive share difficulty you need to add a ‘-d A’ to the start command.
The miner login is then just the payout address without any additions.

Keep in mind, that the rpcuser/password must be the last parameter of the command line ;)

gwbraa

@Wellenreiter said:

@gwbraa I have implemented ‘adaptive share difficulty’ on my p2pool server and just testing it.

It basically implements a different rule set to calculate the share difficulty on a per worker basis.

if you want to try : https://github.com/wellenreiter01/p2pool-neoscrypt
use the ‘adaptive’ branch, not the master branch.

In order to enable the adaptive share difficulty you need to add a ‘-d A’ to the start command.
The miner login is then just the payout address without any additions.

Keep in mind, that the rpcuser/password must be the last parameter of the command line ;)

Just made the chances.
I think now looks like everything it’s happening again!!

Thanks

gwbraa

This post is deleted!

gwbraa

Wellenreite Can you tell me How Can I not connect to the other nodes?
I will set upt a dedicated server and I wanna try run the 200MH/S for some couple hours and I do not want to get everyone not making shares because I am to fast.
Also How can I setup a hostname for the p2pool server??

Thanks

Wellenreiter

Not sure what you mean, your miners cannot connect to other p2pool nodes, or yoir node cannot connect to other p2pool nodes?

In order to set up a nodename for your node, that is known to the Internert, you need to register a domain and then create a DNS Record for the host. There is a small monthly fee for the domain.

With your bandwidth of 200 Mhash, you always will find block so fast, that other nodes have difficulty to submit valid shares before your miner finds a new block.

A possible solution:

go for solo mining with most of your hash power and only keep a hashrate comparable to the hasrate other miners provide or may be double it. This will increase the diffculty of the blockchain, but the p2pool will keep the share difficulty low, so others can submit shares to the p2pool nodes.

wrapper

@gwbraa said:

Wellenreite Can you tell me How Can I not connect to the other nodes?
I will set upt a dedicated server and I wanna try run the 200MH/S for some couple hours and I do not want to get everyone not making shares because I am to fast.
Also How can I setup a hostname for the p2pool server??

Thanks

Re causing large swings in the p2pool difficulty :

I don’t think there will be a problem (at that scale). I received shares for a much lower proportion of the p2pool with both Bitcoin and Litecoin mining.

We have also monitored 10 - 50 MHash/s variations in p2pool, which the p2pool algorithm handled no problem …

In practice what will happen with p2pool if the hash rate goes up, the Global pool members will get more payouts, but a lower rate. When there are few miners / hash, it pays out less often, but higher … The p2pool share system is quite effective at adjusting as the global hash rate changes.

plus we can always monitor the situation, if you let us know when you are testing…

It is theoretically (and practically) possible to make different “colours” of p2pool pools. It is possible to adjust some parameters, so that the p2pool will only connect to a similar compiled pool. This means it would be possible to gave a p2pool adjusted for lots of small hash miners, with lower shares more often, and different coloured pool optimised for high hashing rates, but say short periods (like weekends).

Unfortunately, that sort of facility is not built into p2pool yet, although you can experiment, as it’s open source.

gwbraa

@Wellenreiter said:

Not sure what you mean, your miners cannot connect to other p2pool nodes, or yoir node cannot connect to other p2pool nodes?

In order to set up a nodename for your node, that is known to the Internert, you need to register a domain and then create a DNS Record for the host. There is a small monthly fee for the domain.

With your bandwidth of 200 Mhash, you always will find block so fast, that other nodes have difficulty to submit valid shares before your miner finds a new block.

A possible solution:

go for solo mining with most of your hash power and only keep a hashrate comparable to the hasrate other miners provide or may be double it. This will increase the diffculty of the blockchain, but the p2pool will keep the share difficulty low, so others can submit shares to the p2pool nodes.

I was thinking it was possible to config my p2pool to do not connect to the other nodes, so I would be mining like solo???

About the hostname I already have domains. I already setup one for the machine but I was wondering with in the p2pool I would be able to set the name for the hostname as well.

gwbraa

@wrapper said:

@gwbraa said:

Wellenreite Can you tell me How Can I not connect to the other nodes?
I will set upt a dedicated server and I wanna try run the 200MH/S for some couple hours and I do not want to get everyone not making shares because I am to fast.
Also How can I setup a hostname for the p2pool server??

Thanks

Re causing large swings in the p2pool difficulty :

I don’t think there will be a problem (at that scale). I received shares for a much lower proportion of the p2pool with both Bitcoin and Litecoin mining.

We have also monitored 10 - 50 MHash/s variations in p2pool, which the p2pool algorithm handled no problem …

In practice what will happen with p2pool if the hash rate goes up, the Global pool members will get more payouts, but a lower rate. When there are few miners / hash, it pays out less often, but higher … The p2pool share system is quite effective at adjusting as the global hash rate changes.

plus we can always monitor the situation, if you let us know when you are testing…

It is theoretically (and practically) possible to make different “colours” of p2pool pools. It is possible to adjust some parameters, so that the p2pool will only connect to a similar compiled pool. This means it would be possible to gave a p2pool adjusted for lots of small hash miners, with lower shares more often, and different coloured pool optimised for high hashing rates, but say short periods (like weekends).

Unfortunately, that sort of facility is not built into p2pool yet, although you can experiment, as it’s open source.

Yeah around this number I did not had any problem

wrapper

You are correct, you can mine “solo” - even if your p2pool node does not find others pools, it is connected to the Feathercoin blockchain via a different port.
We have had the situation where p2pool nodes loose contact and the pool continues alone - so you could iptables/firewall, the known other node’s IP address, to deliberately prevent connection to a global pool.

you can use the -getpeerinfo switch with feathercoind to see who’s connected.

So you can find shares, then feathercoind will announce when you found a block and your p2pool will pay you.

Wellenreiter

@wrapper the problem occurs, when blocks are found by the pool so fast, that small miners can’t submit a valid share, before a work restart is requested by the pool.

Yesterday, we had the situation, that blocks where found by the pool about every 10-30 seconds, while miners with hashrates of ~100-150 khash only could send valid shares only every 3-4 Minutes.
The result was, that small miners didn’t get any payout.
This situation is temporary, of course, as the block chain difficulty will go up until blocks are found every minute again, but still such a situation is not encouraging small miners to mine on a p2pool.

RIPPEDDRAGON

said:

t connect to other p2pool nodes, or yoir node cannot connect t

I dont know if you have tried manually setting your miner’s diff but this seems to help me out on p2pool:
-o stratum+tcp://104.236.34.9:19327 -O 6e3ozu798fKvNUCSwRjPMRdnrJ27dCHEPm/0+0.00300:lol

The “0+0.00300” seems to do a decent job of lowering stale % for me

Wellenreiter

@wrapper said:

You are correct, you can mine “solo” - even if your p2pool node does not find others pools, it is connected to the Feathercoin blockchain via a different port.
We have had the situation where p2pool nodes loose contact and the pool continues alone - so you could iptables/firewall, the known other node’s IP address, to deliberately prevent connection to a global pool.

you can use the -getpeerinfo switch with feathercoind to see who’s connected.

So you can find shares, then feathercoind will announce when you found a block and your p2pool will pay you.

The p2pool node initially needs at least one other p2pool node to start.
When this initial connection is lost later, it may continue to work.