schedule179
Posts: 3
Joined: Sat Dec 01, 2018 11:15 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sat Dec 01, 2018 11:31 pm

Re: 557870: Thanks for the heads up. I had forgotten that I had been doing some bitcoind development on server computer while my hashrate was off on pool.bitcoin.com, and was running buggy incomplete code. I fixed this on Wednesday shortly after that orphan block was found.

On another note: Someone with the address 18Wc9XdyCBq8fazP5yErKGA64JVEFENJ1p decided to send about 5 PH/s to my BCH node suddenly. The sudden increase in the number of connections saturated a per-login 1024 open file limit on that account, which made the server get all buggy and drop connections. I've increased the open file limit, and performance is again normal. You're welcome to continue mining there if you'd like.

Warning: i'm seeing quite a few lines like this:

2018-12-01 08:28:43.072649 GOT SHARE! 18Wc9XdyCBq8fazP5yErKGA64JVEFENJ1p 6654496d prev 191c7d81 age 5.77s DEAD ON ARRIVAL

which suggests poor performance (high latency?) on your end. It looks like you're losing about 2% of your hashrate due to DOAs. That's not awful, but maybe worth optimizing.

@jtoomim: That is me. Hopefully it is up to ~7-8 PH/s now. Spinning up my own p2pool node and figured I'd try yours out while waiting for it to sync since we're neighbors in Grant County. How delighted I was to find out you not only contribute/maintain your p2pool git, but that you're also mining BCH on your own node.

Re: the DOA shares, hmm. Absolutely worth optimizing, IMO. Assuming your server is hosted locally I should have fantastic latency to you from only a few miles away. I know my GPU-based private pool server is hosted at a datacenter in Seattle and I get ~5ms ping (from the rig itself to the config server/pool server in Seattle, fully end-to-end). I'm not sure why, I will swap over to my node and see if I am getting the same results and keep yours as failover if you don't mind. Appreciate the work you're doing here. Now I wish I had been in town to come check out your farm open-house after one of the last PUD meetings :)

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Mon Dec 03, 2018 9:05 am

Yeah, my BCH p2pool server is hosted in Moses Lake. Looks like your server is up now, and pool-wide average orphan rates are way down.

Your username is amusing, btw. Is that one of your main financial motivations for mining?

I'm looking back through your shares, and it seems like all of your share difficulties were set to use the minimum allowed while you were mining on my node. The shares that I mined on the same node, on the other hand, have a difficulty that was 30x higher. Were you manually setting a difficulty with the bchaddress/xxxx option? It looks like you're no longer doing that on your own node, and have share diffs at 30x the minimum. This might be the explanation for the performance issues and DOA/orphan rates that you saw for your shares. Low difficulty means short intervals between shares, which means higher orphans and DOAs.

schedule179
Posts: 3
Joined: Sat Dec 01, 2018 11:15 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Dec 04, 2018 4:37 am

Very nice. Building is in Quincy, so should be fine. The original conf did have bch/xxx for diff which I caught and changed after I saw the node crashing, but I don't think I updated the entire farm immediately. Testing some new in-house software and other tweaks, so I didn't notice as quickly as I should have.

Lol, no it isn't a main motivation--I'm just a fan of the original premise of crypto, started off mining in my dorm room back in ~2009 on gaming pc. The 179 stuff was just a huge boost this year and I figure a select few miners will absolutely get it.

kr1z1s
Posts: 5
Joined: Mon Apr 09, 2018 9:23 am

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sat Dec 08, 2018 8:27 am

Yeah, my BCH p2pool server is hosted in Moses Lake. Looks like your server is up now, and pool-wide average orphan rates are way down.

Your username is amusing, btw. Is that one of your main financial motivations for mining?

I'm looking back through your shares, and it seems like all of your share difficulties were set to use the minimum allowed while you were mining on my node. The shares that I mined on the same node, on the other hand, have a difficulty that was 30x higher. Were you manually setting a difficulty with the bchaddress/xxxx option? It looks like you're no longer doing that on your own node, and have share diffs at 30x the minimum. This might be the explanation for the performance issues and DOA/orphan rates that you saw for your shares. Low difficulty means short intervals between shares, which means higher orphans and DOAs.
Hi!

I wrote you here - https://github.com/jtoomim/p2pool/issue ... -445437811

Please, give me you e-mail (in PM) and i send my code to you.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Dec 18, 2018 12:28 pm

I've added dfoderick's changeset for enabling AsicBoost to https://github.com/jtoomim/p2pool/commits/with_df_ab. It seems to work okay on BCH testnet, so I'm testing it now on BCH mainnet. Assuming it works, once it has mined a few blocks, I'll merge it into 1mb_segwit.

Note: if anyone on BCH is getting unusual "naughty share" messages for no apparent reason, this is likely due to a bug in Bitcoin ABC that was fixed in version 0.18.4. Please upgrade to the latest version of Bitcoin ABC to avoid this issue.

https://github.com/Bitcoin-ABC/bitcoin- ... e-notes.md


jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Dec 25, 2018 1:04 pm

Two more AsicBoost BCH blocks have been mined:

https://bch.btc.com/0000000000000000014 ... 9020d132ec - version 0x20400000
https://bch.btc.com/0000000000000000021 ... e315b84af5 - version 0x20000000

The second one was mined with AsicBoost, but ended up with the same version as a non-AsicBoost block. This is expected to happen 25% of the time.

I'm happy enough with these results. I have merged the AsiBoost code into 1mb_segwit.

I've also added a few other changes at the same time, including a performance improvement and some bugfixes when starting new solo-mining chains (e.g. on testnet).

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Jan 15, 2019 10:59 pm

Two BTC blocks were mined on Monday. Congrats, p2poolers!

But BTC p2pool currently has an estimated 33% DOA rate! This is extremely high, and indicates a technical problem for one or more of the large miners. My guess is that one of the large miners is using a large number of unique worker usernames on a single node. P2pool can cache and reuse a single stratum job for many workers as long as their payout addresses and (possibly) worker names are identical, but if the workers are different, new jobs need to be created for each one. Job creation takes around 200 to 800 ms per job on BTC right now, so having more than 5 or so unique workers on a single node can bog down p2pool and cause DOA shares, which results in lower payouts for users of that node. Don't overload your nodes, and try to keep a small number of workers on each node. It's totally fine to have 1,000 S9s on a single node as long as they're all using the same worker name.

rldleblanc is working on code which should improve performance dramatically for p2pool, and should allow more workers per node without this problem cropping up. However, until it's ready, we will all need to be careful to keep the number of unique workers per node low.

Also worth noting: the two recent p2pool BTC blocks were not using AsicBoost.

User avatar
PauloTavares
Posts: 3
Joined: Tue Sep 04, 2018 1:14 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Thu Feb 14, 2019 5:07 pm

Hi
Does anyone knows why Btc p2pool branch as so low hashrate ?

User avatar
bucky911
Posts: 2
Joined: Sat Feb 23, 2019 11:37 pm

Donate BTC of your choice to 3L8DHxgXpzE6SPJRmbzNHxiJKcwfEBW7ag

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sat Feb 23, 2019 11:47 pm

Hi,

I have a couple of newbie questions.

How do I make sure I'm mining on the jtoomimnet BTC p2pool and not the mainnet p2pool?

I read in an earlier post to set blockmaxweight=4000000 in bitcoin.conf, is that still recommended?

Thanks

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sun Feb 24, 2019 9:21 am

If you run code from https://github.com/jtoomim/p2pool/ branch 1mb_segwit, you're on jtoomimnet.

Yes, setting blockmaxweight=4000000 is recommended, as it will allow you to make full blocks.

User avatar
bucky911
Posts: 2
Joined: Sat Feb 23, 2019 11:37 pm

Donate BTC of your choice to 3L8DHxgXpzE6SPJRmbzNHxiJKcwfEBW7ag

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Mar 05, 2019 2:21 pm

I've modified my local p2pool to work with Bitcoin v0.17

If the default max block weight in v0.17 makes full blocks, I think I should remove blockmaxweight=4000000 from bitcoin.conf

Could this cause issues with my peer p2pool nodes?

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sun Mar 31, 2019 11:11 pm

No, a p2pool node can set blockmaxweight however it wishes without affecting other nodes on p2pool.

OneChain
Silver Bitcoiner
Silver Bitcoiner
Posts: 1086
Joined: Sat Aug 19, 2017 10:11 am
Contact: Telegram

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sat Apr 20, 2019 12:46 pm

I'm curious, what was the reason behind the creation of a "vip" btc p2pool?
Whichever chain wins,the world wins Image

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Jun 18, 2019 5:02 am

@OneChain it was because there was a bug with the Antminer S9 firmware that choked as soon as the coinbase transaction reached 8 kB, which happened at around 130 addresses IIRC. Some people wanted to reduce the number of users and payout addresses, so they created their own forked sharechain. Modern S9 firmware does not have this bug.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Mon Jul 01, 2019 7:14 am

A major upgrade has been merged into https://github.com/jtoomim/p2pool (master branch). This upgrade brings many major new features, thanks to rldleblanc:

1. Segwit payout address support on BTC and LTC
2. P2SH support
3. Cashaddr support on BCH
4. Huuuuuuuge performance improvements and stale/orphan/DOA rate reductions

This release also comes with a small cost:

5. P2pool's built-in fast-block propagation algorithm was removed, as it was obsolete (https://github.com/jtoomim/p2pool/issues/19).

Unlike previous releases, this is in the master branch. If you are already using my repository, you can upgrade like this:

> git checkout master
> git pull

The installation instructions in the README.md (at the bottom of https://github.com/jtoomim/p2pool) have also been updated to reflect the usage of master branch. If you're creating a p2pool installation from scratch, they should work for you.

This upgrade is a hard fork. The new features will only take effect once the new version (v34) has been activated. Once 50% of the hashrate is signaling for v34 reaches 50%, a warning message will be displayed to all existing p2pool users. Once 95% of the hashrate is signaling for v34, the upgrade will be activated and nodes will start creating v34 shares, and the new features will be available.

I will be delaying the upgrade of my own BCH and LTC nodes for one week to give other people a chance to upgrade first and to avoid having the fork be triggered prematurely on those networks.

If you choose not to upgrade and the 95% threshold is reached, your node and all other non-upgraded nodes will be forked off the network. They will continue to work normally, but without the help of everyone else's hashrate. The 1mb_segwit branch is being left intact for v33 in case anyone wishes to downgrade or otherwise remain on v33.

User avatar
rogerver
Founder
Founder
Posts: 1868
Joined: Thu Sep 10, 2015 6:55 am

Donate BTC of your choice to 1PpmSbUghyhgbzsDevqv1cxxx8cB2kZCdP

Contact: Website Twitter

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Mon Jul 01, 2019 5:36 pm

Thank you for your great work and follow up over the years Jonathan!
Help spread Bitcoin by linking to everything mentioned here:
topic7039.html

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Jul 02, 2019 3:03 am

I pushed a few more changes to github master today. I recommend doing a git pull if you're mining BCH. These changes allow for much larger blocks to be mined with P2pool. Once mainnet forks, P2pool will be capable of mining 31.999 MB BCH blocks with tolerable performance on a fast (high GHz) machine:

https://explorer.bitcoin.com/tbch/block/00000000000001f6b6068c3fc228c9d1388d7358515a997d2a1cf8a644c106ac

Most p2pool users will want to limit their blocks to 8 MB most of the time. But that's still a 4x improvement over before, where even 2 MB was pushing p2pool to its limits.

More optimizations are underway, and I hope that soon 32 MB blocks will be easy to mine on any desktop computer.

frstrtr
Posts: 3
Joined: Wed Jul 03, 2019 12:49 pm

Donate BTC of your choice to 1KPhVLJKFpfSfKw3hntFvXrZvreoaxfi99

Contact: Telegram

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Wed Jul 03, 2019 12:53 pm

Installing a new node from Master on Ubuntu 18.04.2LTS Server for BTC...

got some errors like
"
p2pool.util.jsonrpc.NarrowError: -8 hash must be of length 64 (not 45, for '4105aec3a75c5f3f7daf737f2481fc54951b758151538')
"
Any suggestions?
Thanks.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Mon Jul 08, 2019 3:26 am

The "hash must be of length 64" error has been fixed. Do a git pull.


jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Fri Jul 12, 2019 9:07 am

The BCH and LTC p2pools have both forked to v34 earlier today. Everything appears to be working so far. BCH has mined two post-fork blocks successfully. LTC hasn't had any blocks yet, but no major problems have been observed.

Old v33 nodes are still connecting to v34 nodes, requesting all of the v34 shares, finding them invalid, disconnecting, then reconnecting and repeating. This is wasting a bit of bandwidth. This could have been avoided if we had incremented the protocol version to 3401, but I forgot to do that and rldleblanc (who wrote most of the code this time) probably didn't know any better. Oh well. Because the v34 shares are so super-duper tiny now, it's not using very much bandwidth -- maybe 5 kB/s extra on average -- so I'm not concerned about it, and might not bother to fix it. As the remaining v33 nodes upgrade, this problem will likely go away on its own. If it doesn't, I may add peer tempbanning for cases of repeated problems like this.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Tue Jul 16, 2019 3:05 pm

I've added a couple of beta code branches:

34-memfree:
1. Better stratum job caching -- no longer creates new stratum jobs for different usernames that resolve to the same address. This reduces CPU usage and DOA rates for some nodes, and was necessary to allow many differently-named workers on a single address.
2. Contains some other changes which might reduce memory consumption and CPU usage, but probably don't do anything
3. Allows the use of username.worker notation, or username_worker notation. username can be a mining address, or it can be any other string (for mining with the node's default address).
4. Preliminary web UI support for graphing multiple workers. Not yet pretty, but sorta functional. You may need to shift-reload the graph page, otherwise your web browser will keep the cached static HTML that can't view workers well. Visit http://ml.toom.im:9348/static/graphs.html?Week and ctr-f for tbs9-200 for an example of the new graph UI.
5. Will temporarily ban peers that disconnect suddenly while processing shares (e.g. nodes running old code).

stratum-autodiff:
1. Everything in 34-memfree
2. Automatic adjustment of stratum job difficulty for each stratum connection independently
3. New command-line option: -s x or --share-rate=x to set the targeted number of seconds per share per connection. Default: 3.0. Set this to a very low value (like 1.0 or 0.1) for nodes with very few LAN miners for smooth hashrate graphs. Set to a high value (10 or higher) for large nodes with hundreds or thousands of miners to avoid CPU saturation.
4. Much cleaner statistics graphs using the new default configuration.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Fri Aug 02, 2019 7:22 am

The BTC p2pool forked to version 34 about 1 week ago, on or about July 23rd. Orphan rates have dropped from about 4% to about 0.9%. DOA rates have fallen as well, from 11% to 3.8%. Overall stale share rates dropped by about 3x, and good rate increased from about 85% to 96%.

Network traffic also dropped dramatically. Before the fork, my p2pool nodes were using about 100 kB/s each, with about 50% upload and 50% download. After the fork, my nodes were using about 8 kB/s each, and almost all of that was upload, due to v33 nodes repeatedly connecting to v34 nodes and downloading the recent shares over and over. After upgrading my nodes to the stratum-autodiff branch, average bandwidth fell to 1.14 kB/s.

Overall, a very good improvement.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Wed Aug 28, 2019 6:37 am

Toomim Bros has shut down its 350 GH/s of LTC p2pool hashrate. We expect this change to be permanent.

We expect to add around 8-9 PH/s of SHA256 hashrate to p2pool soon to replace it. This hashrate will probably be mining BCH, but there's a chance I might do BTC with it instead.

frstrtr
Posts: 3
Joined: Wed Jul 03, 2019 12:49 pm

Donate BTC of your choice to 1KPhVLJKFpfSfKw3hntFvXrZvreoaxfi99

Contact: Telegram

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Wed Aug 28, 2019 8:42 am

Can i still use toom_im 9327 as a backup pool-node for my miners?

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Wed Aug 28, 2019 9:10 am

I'll leave up ml.toom.im:9327 for now. I might eventually take it down if the RAM consumption gets to be a problem, but I'll try to track you down and let you know. Maybe send me your email address? j at toom dot im

schedule179
Posts: 3
Joined: Sat Dec 01, 2018 11:15 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Mon Oct 28, 2019 10:26 pm

I'll leave up for now. I might eventually take it down if the RAM consumption gets to be a problem, but I'll try to track you down and let you know. Maybe send me your email address?
Just checking in to say thanks for all the work you do with p2pool (specifically for BCH-friendly miners like myself). I would be interested in helping support the work, if possible, in some fashion.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Sat Feb 08, 2020 6:05 am

We recently encountered a bug on the BCH p2pool. 1/256th of all user addresses will be incorrectly translated into output scripts, and cannot be paid by p2pool. The bug comes from addresses for which the HASH160(pubkey) begins with a leading zero (0x00) -- in this situation, p2pool will submit a script with a pubkeyhash length of 152 bits instead of 160 bits. This prevents the funds from being spendable.

In order to fix this bug, a hard fork of the p2pool sharechain is required. All nodes in the p2pool network need to be able to agree on the contents of the coinbase transaction at any given point in the share chain. Fixing this bug on some nodes but not others would result in a chainsplit.

rldleblanc and I have written hardfork code to fix the bug. I will be releasing it soon. Upgrading on BCH is urgent, as any blocks created before this fork is activated will currently contain a fairly large (roughly 9% of the block reward) unspendable output (unless the sharechain gets purged of the shares currently belonging to the user whose address triggers the bug). Consequently, we will be doing a rushed hard fork. I will probably trigger the BCH upgrade as soon as 60% of shares over the last day supports the fork, rather than the customary 95% threshold over 3 days -- that is, as soon as possible while maintaining network consensus. Other coins should upgrade soon, but less urgently, and can wait for 95% support.

Keep an eye out for the upgrade.

jtoomim
Nickel Bitcoiner
Nickel Bitcoiner
Posts: 135
Joined: Fri Nov 06, 2015 7:46 pm

Re: P2pool: Decentralized, DoS-resistant, trustless, censorship-resistant pool.

Mon Feb 10, 2020 1:22 am

Update required. The p2pool hard fork changes to fix the padding bug have been merged into master. Please update your nodes by doing git checkout master and git pull. I will be moving my hashrate back onto main p2pool later today, and expect the fork to activate within 3 days.

https://github.com/jtoomim/p2pool/commits/master

Return to “Pools”

Who is online

Users browsing this forum: No registered users and 17 guests