[1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool

[twmz]

Seems like SQL Server eating up a lot of memory due to historical data?
That number of visitors should not be a problem for a medium sized server.

It's not the SQL Server that is the problem.  It's the web server.  The site does a lot of number crunching every time it calculates block durations, luck, etc.  That's CPU expensive, so it also caches essentially all of the results in memory (which include, among other things, hashrate data points for every 5 minutes for the past year+, so lots and lots of data in RAM on the web server).  I imagine there are other inefficiencies as well that are in the code, but I have not taken the time to investigate and track them down.  You're welcome to do so once the code is on GitHub.

It would fit fine on a server if it was by itself, but it's not.  It's on the same virtual machine as my other site with is much more popular and a more intensive web application.  There's just not room for the p2pool app on that VM anymore.  If the p2pool app used minimal resources, it would be fine, but it doesn't.

[ok]

Seems like SQL Server eating up a lot of memory due to historical data?
That number of visitors should not be a problem for a medium sized server.

It's not the SQL Server that is the problem.  It's the web server.  The site does a lot of number crunching every time it calculates block durations, luck, etc.  That's CPU expensive, so it also caches essentially all of the results in memory (which include, among other things, hashrate data points for every 5 minutes for the past year+, so lots and lots of data in RAM on the web server).  I imagine there are other inefficiencies as well that are in the code, but I have not taken the time to investigate and track them down.  You're welcome to do so once the code is on GitHub.

It would fit fine on a server if it was by itself, but it's not.  It's on the same virtual machine as my other site with is much more popular and a more intensive web application.  There's just not room for the p2pool app on that VM anymore.  If the p2pool app used minimal resources, it would be fine, but it doesn't.

I see, will take a look once the code is on GitHub.

[daemondazz]

Is zvs the developer of the p2pool?

forrestv is.  

Ah, thx.

Did anybody know, how can i setup p2pool to generate p2pool-share from all miners globally?
And pay them out over there shares on my node?

I know its not "hop-proof" but i have many small miners with usb sticks, and the p2pool-diff is too high for a single share from a miner.
So i mean many miners works on a node-share to p2pool network and become payouts based on miner shares to the node.

Or must i use eloipool ?

Have a look at cryptominer.org - this is my node that is running a script that does something similar to that (miners are paid for their percentage of work done on my node). My modifications to the p2pool codebase are at https://github.com/forrestv/p2pool/pull/127

[forrestv]

The Future of p2pool.info

As some of you know, I developed the p2pool.info website back when p2pool first started to take off because I was looking for a way to understand the blocks that p2pool finds, the pool's hashrate over time, etc.  This was at a time before p2pool has its own local web interface built in.

I no longer mine on p2pool info myself, but I've been happy to continue hosting p2pool info for the benefit of those that do because it currently runs on the same web server as one of my other non-bitcoin related project, and I'm already paying for those servers anyway.

Thank you, twmz, for the long time you've spent supporting p2pool.info and P2Pool. Having p2pool.info around undoubtedly helped P2Pool become as popular as it is now.

P2Pool's built-in web interface definitely doesn't satisfy the same needs that p2pool.info does, and so I'd like to take on maintenance of p2pool.info. I'll bring up a suitable host and then I'd like to take ownership of the domain name and move it to that host for now, and then probably in the long term rewrite it to work on a Linux VPS.

[Polyatomic]

fully blown sickness forrestv.

[twmz]

The Future of p2pool.info

As some of you know, I developed the p2pool.info website back when p2pool first started to take off because I was looking for a way to understand the blocks that p2pool finds, the pool's hashrate over time, etc.  This was at a time before p2pool has its own local web interface built in.

I no longer mine on p2pool info myself, but I've been happy to continue hosting p2pool info for the benefit of those that do because it currently runs on the same web server as one of my other non-bitcoin related project, and I'm already paying for those servers anyway.

Thank you, twmz, for the long time you've spent supporting p2pool.info and P2Pool. Having p2pool.info around undoubtedly helped P2Pool become as popular as it is now.

P2Pool's built-in web interface definitely doesn't satisfy the same needs that p2pool.info does, and so I'd like to take on maintenance of p2pool.info. I'll bring up a suitable host and then I'd like to take ownership of the domain name and move it to that host for now, and then probably in the long term rewrite it to work on a Linux VPS.

Awesome.  I'll find you on IRC sometime or send you a PM to discuss the details and to help get it up and running elsewhere.

[oldbushie]

Thank you for taking up the mantle, forrestv! p2pool.info is incredibly useful.

Speaking of VPS, does anyone have recommendations for running a p2pool node on a VPS? Any reliable hosts that are reasonably priced? I have no clue what I should be looking for in terms of minimum RAM and disk space and bandwidth.

I'm currently leaning towards myhosting.com since their plans are fully customizable.

[phillipsjk]

Thank you for taking up the mantle, forrestv! p2pool.info is incredibly useful.

Speaking of VPS, does anyone have recommendations for running a p2pool node on a VPS? Any reliable hosts that are reasonably priced? I have no clue what I should be looking for in terms of minimum RAM and disk space and bandwidth.

I'm currently leaning towards myhosting.com since their plans are fully customizable.

At home, Bitcoin and friends are getting a dedicated machine due to expected disk I/O. While activity should drop once the chain is caught up, I doubt many VPS hosts will be happy with that type of usage.
 

[oldbushie]

I ended up going with myhosting.com with 100GB disk space, 4GB RAM, and 300GB monthly bandwidth (for now). The plan is easily expandable so I can always bump up the bandwidth, for instance. It's all on SSDs so disk I/O shouldn't be an issue. Syncing the blockhain overnight and then I'll try installing p2pool on it.

[ok]

The Future of p2pool.info

As some of you know, I developed the p2pool.info website back when p2pool first started to take off because I was looking for a way to understand the blocks that p2pool finds, the pool's hashrate over time, etc.  This was at a time before p2pool has its own local web interface built in.

I no longer mine on p2pool info myself, but I've been happy to continue hosting p2pool info for the benefit of those that do because it currently runs on the same web server as one of my other non-bitcoin related project, and I'm already paying for those servers anyway.

Thank you, twmz, for the long time you've spent supporting p2pool.info and P2Pool. Having p2pool.info around undoubtedly helped P2Pool become as popular as it is now.

P2Pool's built-in web interface definitely doesn't satisfy the same needs that p2pool.info does, and so I'd like to take on maintenance of p2pool.info. I'll bring up a suitable host and then I'd like to take ownership of the domain name and move it to that host for now, and then probably in the long term rewrite it to work on a Linux VPS.

Awesome.  I'll find you on IRC sometime or send you a PM to discuss the details and to help get it up and running elsewhere.

Nice to hear that p2pool.info will continue. It's a good source for historical data.

[zvs]

I ended up going with myhosting.com with 100GB disk space, 4GB RAM, and 300GB monthly bandwidth (for now). The plan is easily expandable so I can always bump up the bandwidth, for instance. It's all on SSDs so disk I/O shouldn't be an issue. Syncing the blockhain overnight and then I'll try installing p2pool on it.

ovh.com has some machines in Canada that aren't sold out anymore.  The $29 one would be plenty..

or, you could try kimsufi.fr, ~$13 if you don't live in the EU.  Not sure if that processor could handle it, though

ed: https://www.datashack.net/dedicated/ is a budget provider in the US (missouri)

[oldbushie]

Alright, the node seems to run fine on myhosting. Now to run it overnight and watch for any hiccups.

[zvs]

300gb bandwidth is ... nothing

i hope it doesn't charge much for overage

[Red Emerald]

I stopped mining awhile ago, but I've left my p2pool node up.  It's averaged 36 peers in the last year and it's bitcoind node is well connected.  It has a high speed internet connection and has had excellent uptime.

Is a node like mine actually helpful for p2pool?  I'm assuming that having a well connected node is good for spreading transactions and protecting p2pool from some kind of network attack.  Am I right, or is a non-mining node not helpful? It's just a git pull and a restart ever now and then, so it's not much effort.

If a non-mining node is helpful, is there anything else I can do to make it more useful?  Maybe set the max connections to something a lot higher so the pool sees more transactions?  I'm just guessing.

[Carlton Banks]

I stopped mining awhile ago, but I've left my p2pool node up.  It's averaged 36 peers in the last year and it's bitcoind node is well connected.  It has a high speed internet connection and has had excellent uptime.

Is a node like mine actually helpful for p2pool?  I'm assuming that having a well connected node is good for spreading transactions and protecting p2pool from some kind of network attack.  Am I right, or is a non-mining node not helpful? It's just a git pull and a restart ever now and then, so it's not much effort.

If a non-mining node is helpful, is there anything else I can do to make it more useful?  Maybe set the max connections to something a lot higher so the pool sees more transactions?  I'm just guessing.

More importantly, your node will help with propagating blocks found by p2pool users. The more pervasive the p2pool network is, the fewer blocks are likely to get orphaned. forrestv apparently has something in the p2pool code that distributes the block super fast to prevent as many orphans, but I think that more nodes will also help to do the job too.

p2pool is unique in this respect, I don't think any other pool has a dedicated network of relay nodes running code to help propagate blocks, the centralised pools probably have to keep an eye on who their bitcoind nodes are connecting to and make sure they are using the most reliable nodes, the most well placed and those with the most capacious bandwidth.

[lanfanblue]

Hi forrestv and all contributors of p2pool,

Many thanks to this fantastic mining pool. It's really fun to tuning p2pool to maximize miner's income for geeky users like me.

But I wonder if you have any plan to enable IPv6 support of p2pool?
Since official bitcoin client has IPv6 enabled by default, I think enabling IPv6 support will make p2pool network even more pervasive and well connected.
More importantly, IPv6 will help those users with both v4 & v6 network, but behind firewall or NAT in v4 network, just like me.
I do have a very good connectivity to v6 network and can accept income connections.
But v4 network is NATed with high latency, and I think that might be the source of my high orphan rate.

I have checked that twisted library used in p2pool support IPv6 network, at least partially.
I'm not familiar with python, so I don't know for sure.
I wonder if it is possible to enable IPv6 network in p2pool?

Thanks again to this great mining pool. :-)

[twmz]

The Future of p2pool.info

As some of you know, I developed the p2pool.info website back when p2pool first started to take off because I was looking for a way to understand the blocks that p2pool finds, the pool's hashrate over time, etc.  This was at a time before p2pool has its own local web interface built in.

I no longer mine on p2pool info myself, but I've been happy to continue hosting p2pool info for the benefit of those that do because it currently runs on the same web server as one of my other non-bitcoin related project, and I'm already paying for those servers anyway.

Thank you, twmz, for the long time you've spent supporting p2pool.info and P2Pool. Having p2pool.info around undoubtedly helped P2Pool become as popular as it is now.

P2Pool's built-in web interface definitely doesn't satisfy the same needs that p2pool.info does, and so I'd like to take on maintenance of p2pool.info. I'll bring up a suitable host and then I'd like to take ownership of the domain name and move it to that host for now, and then probably in the long term rewrite it to work on a Linux VPS.

The source code is now on GitHub:  https://github.com/ervwalter/p2pool.info

I'll work with forrestv to get it up and running on a new official server and then will transfer the domain name to him.  It's open source and so other are welcome to setup their own unofficial copies of the site, if they want to play with it for some reason.  Out of the box, the easiest (but not the cheapest) way to deploy it is on Windows Azure.  It's as simple as setting up a web server and a database server there and then pushing the code to the web server using git.  Setting it up on a non-Azure windows server is probably relatively straightforward as well (and may be much cheaper), but the specifics are left as an exercise for the reader.

[Exidous]

The source code is now on GitHub:  https://github.com/ervwalter/p2pool.info

I'll work with forrestv to get it up and running on a new official server and then will transfer the domain name to him.  It's open source and so other are welcome to setup their own unofficial copies of the site, if they want to play with it for some reason.  Out of the box, the easiest (but not the cheapest) way to deploy it is on Windows Azure.  It's as simple as setting up a web server and a database server there and then pushing the code to the web server using git.  Setting it up on a non-Azure windows server is probably relatively straightforward as well (and may be much cheaper), but the specifics are left as an exercise for the reader.

Any way to convert the .bacpac to sql?

[Red Emerald]

Hi forrestv and all contributors of p2pool,

Many thanks to this fantastic mining pool. It's really fun to tuning p2pool to maximize miner's income for geeky users like me.

But I wonder if you have any plan to enable IPv6 support of p2pool?
Since official bitcoin client has IPv6 enabled by default, I think enabling IPv6 support will make p2pool network even more pervasive and well connected.
More importantly, IPv6 will help those users with both v4 & v6 network, but behind firewall or NAT in v4 network, just like me.
I do have a very good connectivity to v6 network and can accept income connections.
But v4 network is NATed with high latency, and I think that might be the source of my high orphan rate.

I have checked that twisted library used in p2pool support IPv6 network, at least partially.
I'm not familiar with python, so I don't know for sure.
I wonder if it is possible to enable IPv6 network in p2pool?

Thanks again to this great mining pool. :-)

I'm all for IPv6, but will adding IPv6 support be enough for you?  I think you will still likely need to talk to IPv4 peers for a while since I doubt everyone in the pool can make the switch.

[twmz]

The source code is now on GitHub:  https://github.com/ervwalter/p2pool.info

I'll work with forrestv to get it up and running on a new official server and then will transfer the domain name to him.  It's open source and so other are welcome to setup their own unofficial copies of the site, if they want to play with it for some reason.  Out of the box, the easiest (but not the cheapest) way to deploy it is on Windows Azure.  It's as simple as setting up a web server and a database server there and then pushing the code to the web server using git.  Setting it up on a non-Azure windows server is probably relatively straightforward as well (and may be much cheaper), but the specifics are left as an exercise for the reader.

Any way to convert the .bacpac to sql?

Not that I know of.  However, I re-exported the database as .sql files as well.  Note, these are still SQL Server specific.  Also, they are massive (as compared to the binary .bacpac file), so I moved all the data exports to its own repository:

https://github.com/ervwalter/p2pool.info-data