6 hours ago, shares be dropping :)

Is anyone charting the p2pool hashrate?

logs should be enough to chart it later

or blocks found

just look at difficulty over time

from the jobs

but yes it'd be a nice chart to have

Would be fun to see how quickly it's growing/what pools the hashrate is coming from

crypto_grampy[m] minexmr.com/pools has graphs for p2pool hashrate if you zoom in really really far.

Oh, you can also click on it on the legend to select/unselect it

double click for only showing that

sech do we need to roster the

Sech do we need to update our Monerod to the new version or is the old one fine to use?

garth you should still be using the one that you compiled from GitHub because the zmq changes didn't make it into this release.

(Newest one works for wallet and tracking mined outputs, just not as the p2pool monerod)

I'm still using a 4 day old build

Is that fine?

it's better to update to latest, I don't want to catch bugs I already fixed

crypto_grampy minexmr.com/pools

Aaaa it takes 10 minutes download and build required things and me lazy

Also people talking about build errors here

They still present?

It should build on ubunutu, macos and windows

Is it meaningful for low hashrate miners to try p2pool? (10-20kH/s)

Give it a go for a week and check results

:)

I have several payouts a day with 30 kh/s

even 1 kh/s miners will get a payout once every few days

Wait are all crypto exchanges a loss of money for exchanger?

Checking the sites they say 1% fee but end result would be 10% loss

hmm, any reason why clan has these defined? github.com/SChernykh/p2pool/blob/master/src/log.h#L174-L177

it... fails compilation on clang

p2pool/src/log.h:175:1: error: redefinition of 'Entry<long>'

INT_ENTRY(long)

p2pool/src/log.h:168:1: note: previous definition is here

INT_ENTRY(int64_t)

(also the second definition there)

it's from initial push, so can't find any reason why, is it related to a specific clang version?

running clang 10 here

this was for macos

hmm, then maybe ifdef can be targeted a bit better

yes, it looks like #ifdef __APPLE__ works too

replaced it locally to #if defined(__clang__) && defined(__APPLE__)

unless it's apple-specific for gcc and so too

there's only clang on macos

"murr4y: Is it meaningful for low hashrate miners to try p2pool? (10-20kH/s)" My hashrate varies between 2-4 kH/s (depending on whether my son's gaming rig is participating or he's using it) but on average for week 3 Kh/s and I've gotten three payments so far this week.

(it's an old ass gaming rig BTW)

btw sech1 had to specifically apt install libgss-dev

otherwise apparently did not have libgss

(classic /usr/bin/ld: cannot find -lgss)

this was an Ubuntu 20.04 flavor

it's in the build instructions in readme

doh, copied the monerod one

just doing some local testing, was quite confused

maybe should have gotten more sleep

oh, new block

growing, 14 outputs now

sech1 One question for my understanding about the shares in the PPLNS window, do shares get wiped as soon as we find a block or they stay in the window for the whole time. i.e. if I have a share, we find a block, I get paid on that block, then we find another block can I get paid again for same share?

they stay in PPLNS window for 2160 pool blocks

The reason I ask, is I'm looking at show_transfers in my wallet and all my payouts are w/in the same 6 hour window... oh just saw reply... makes sense.

one share can give you many payouts if the pool is lucky

Awesome. That's how I thought it worked. I was just confirming because I had told some people on Reddit that's how it worked.

Wait but doesn't it break if the pool gets more miners and blocks start to be found more often?

Like when I would find 1 share and in 6 hours pool finds 10 blocks

Where the money comes I dont understand

:D

ooooof.. I don't where to begin with that..

I lost 1 out of my 2 braincells trying to understand it

Every block has a block reward and fees. So the payment to those miners is coming from each blocks reward as coinbases same as normal.

Do you understand how pool mining works in general?

wrote long ass explanation but deleted it for being too complicated.. :D

Me? No. I'm just explaining where the payment is coming from. I couldn't explain the philosphy behind PPLNS and why it's "fair".

sorry I meant QuickBASIC

I meant

AAAAAA mightysnowman

I understand normal pool mining

Where I help mine one block and get rewarded for some % of that block

It's excactly the same thing, execpt there is share culling happening

yes and you will get paid for that 1 share within that 2160 ppln window

the % changes if other p2pool miners send more shares or don't send shares

Pools use PPLNS windows or similar as well, this is no different.

You get paid in proportion to your hashes during the PPLNS window, per block found

lets say there are 9 shares submitted within that 2160 ppln window, 1 is yours. block is found by someone, the total shares becomes 10, you get paid for 1 share, 10% of total block value.

If you're 10% of the p2pool hashrate you get 10% of each block found during the window.

later the share amount increases to 19 shares, still 1 is yours, block is found you get 5% of total block value

etc.

if the share is older then 2160 sidechain blocks you won't get paid

Ooh cool makes sense

Only if I'd ever get any shares

What's your hr mightysnowman?

8kh

Perhaps the problem is that I havent mined p2pool 24/7

Only like 6 hours each day

Could it be?

Oh, so you're effective hr is really low if you're not mining all day.

It would be the same as mining with 2kh/s 24 hours.

What is effective hr

Oh okay

Deem I need to get myself separate rig for mining

Its my only pc and cant really use it for work while mining

Why is that?

I need to run resource intensive stuff

ah right

you could try --cpu-priority 0

I can do pretty much everything with all threads in use :)

It isnt by default?

nope

Oh okay makes sense will try lowering priority

Why monero price rn go brr down

60€ drop if I'm looking at the right graph

cause the corn went down

also, #monero-markets

The last time I looked in there they were talking about masking and conspiracy theories. No thanks.

lol

mightysnowman: just a normal that kills leverage, carry on

what are the "not found" during building, are they a problem? irccloud.com/pastebin/kfpXgswD

that is fine, it's just looking at what to use on your system mightysnowman

normal cmake/configure etc behavior

okay

a noob question but can I just combine those 2 things into single .sh script?

like so

it works but made folders in different places for some reason

compared to manual copy/paste

I ran the script from ~ just like the shell by default does

When the second part of the script runs you're in ./p2pool/build/ so it clones the Monero repo there. Try adding a cd ../.. in between the blocks so you go back to where the script is run from.

Also, you only need to install the packages once so you can remove both apt-get lines if they're already installed.

If you run the script again though, you're going to get git errors because the repo folders already exist, so you'll need to use git fetch instead.

^ mightysnowman

git fetch? isn't git pull enough?

didnt get any git-related errors

ran the stuff for the same dir

jaska087 ah yeah probably. I just know you can't use clone if it exists already.

~/.src$ git clone --recursive github.com/SChernykh/p2pool

fatal: destination path 'p2pool' already exists and is not an empty directory.

wait is there a command than could do either clone or pull automatically_

yes, that why i'm a bit concerned he didn't get any errors

?

perhaps there were errors but cant find em because big history

He was probably in a different pwd

yeah it seems so

Just strike line 2 and s/cd p2pool/cd p2pool && git pull/

But don't lend me too much credence... most of my bash knowledge is from SunOS lol.

so that should be?

added the cd and replaced clone with pull

I don't think you can pull from outside the directory like that.

it needs the .git file in the same dir

so just add .git to end_

?

this is the moment where I would recommend some kind of jail or container

what

I googled, it should work tho

pull and then directory

could you try?

would recommend specifically checking if you have already pulled, then fetch and reset to head instead

or... just nuke it :>

yeah

sudo rm /*

-r

lol

ayo my linux gone

DataHoader my assumption is that he likely hasn't made any local changes, so shouldn't have to reset to head.

oh you are doing pull without the clone there?

remote changes can happen

and they can be conflicting QuickBASIC

not just mergeable

for example if force pushes happen

lawd av mercy. git is confusing... I thought git pull was basically shorthand for git fetch then git merge FETCH_HEAD with changes?

which can conflict yes

Okay. So what should I be doing to update my local repo when I recompile, then?

note you can also just wget a tar.gz or zip of the repo

git fetch && git reset --hard origin/master && git submodule sync && git submodule update --init --force --recursive

for monerod afaik

as submodules can also change yay

or get added/removed

Why does the p2ppool wallet-cli show incoming transactions from mining and the normal gui wallet not?

Urgh.. linux is there a way to just dump text into a file... i.e. like TYPE in MS-DOS... TYPE > file.txt, then SHIFT+INS

xmrpow: you might need to upgrade to latest gui

or follow instructions on README

oh, they were removed now

@QuickB echo "hallo" > test.txt

latest GUI shows the special outputs mined

QuickBASIC: cat > file

type things now :)

or paste

Ah thank you.

xmrpow: Official Monero CLI and GUI v0.17.2.3 and newer

if you update to that it should do it

alternatively

> - Open this wallet in CLI: run `./monero-wallet-cli`, enter the wallet file name there and then enter the command `set refresh-type full`. **This step is important!** If you don't do it, you won't see p2pool payouts!

then you can open with GUI again

Ok thank you!

no idea if you might need to rescan the last few blocks to see the outputs again

ill try

also what are 1gb pages?

use v0.17.2.3 and rescan the wallet

what are they for if I have huge pages enabled?

rn I have 1gb pages disabled

better one off allocation mightysnowman, instead of using many small hugepages it uses fewer 1g

ty

smaller performance increase compared to having normal hugepages in the first place

how's the electricity cost for you guys?

kwh

0.062€/kwh here

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA remember how I deleted the monero and p2pool folders before

where is the default wallet file location

luckily it didnt get any payouts yet

always backup your seed D:

more than once

didnt bother since made a new wallet for p2pool as it told me to and since p2pool beta I didnt think I'd keep mining or anything

around 0.12€ kw/h

electricity itself costs about 0.046 kw/h, rest of the price is bulshittery

:)

what does this lock mean and how to remove it

it seems to cause problems

you've created that file with sudo/root rights?

yes

Well there's your problem

how to fix it

there was linux in school but the teacher was a bitch so I dont know anything

mightysnowman: you can use the same wallet in the end, just needed initially a few settings set

for the file/folder mode

check chmod / chgroup

chown*

im losing my brain

nvm typo

I'm too spoiled by windows

linux hurts my brain

I'm too spoiled by linux

windows hurts my brain :)

what is this thing called a "mouse pointer"

windows hurts itself twitter.com/arstechnica/status/1435290023503405057

I mean I would learn linux and use it daily only if I wasn't stuck with windows propiretary software and games and stufff

wine is unable to emulate all of them

hopefully Steam deck makes devs focus on linux more

srsly how can anyone still trust microsoft after all this time, such a clear track record

PoweShell spoils me. Linux terms don't atab complete parameters and it annoys me.

/s/atab/tab

Type 'Get-Mailbox -' tab tab tab to see all the available parameters etc.

For some things trust Microsoft. There's a reason M$ Exchange + Exchange Online has 65% market share.

that's what <cmd> -h (or --help) is for

mightysnowman: proton ftw :)

and QuickBASIC it tab completes here

just need to have proper tab completion added

Linux is the kernel, blame userland :)

hyc: "that's what <cmd> -h (or --help) is for" if I'm in the middle of typing a command I have to open another term or screen or web browser.

nah you can save commands

that's what the terminal is for :)

but yeah it takes time to get used to everything

lol twitter.com/CSA_DVillamizar/status/1432870560774692865

DataHoarder: "Linux is the kernel, blame userland :)" fair enough but coming from a Windows background I expect sane tab complete to be a default for whatever distro's default terminal is installed.

Yeah thankfully that's out of my support scope. Our Azure guys were scrambling though lol.

so much for Exchange...

Yeah, but that's people neglecting to patch. There's been only one Exchange zero day in the wild in the last year, but I admit you can't always trust Microsoft.

it all reflects a trend of growing stupidity. you can't trust them to secure your desktop

I actually only do support for O365 and some hybrid configs. on-prem is a whole other problem to secure.

and now they're pusing all their customer into the cloud, which is even less trustworthy

and the scope of a breach in the cloud tends to dwarf anything that could be breached on-prem

aside from the fact that they no longer make any pretense of you "owning" any of the software. now it's all perpetual lease, pay up and keep paying

they are the very picture of evil capitalism, just rent-seeking trash.

I don't disagree, but they print my paycheck every month on time and I'm not crypto-rich yet.

But it was a mistake to defend them. Exchange is trash, but most businesses DO trust and use it for some reason.

I wonder if UBI would make unethical companies like that go extinct

Maybe. I was sad when Andrew Yang dropped out.

finally a block!

Urgh I think I broke something when I recompiled monerod. Thankfuly my miners jumped over to my NAT'd p2pool

Pushed a Docker image for monerod off of the p2pool branch: sethsimmons/simple-monerod:p2pool-api-v0.17

one more block

yay

talking about docker stuff sethsimmons, pushing a repo out here soon

if you want a docker-compose setup github.com/WeebDataHoarder/p2pool-compose

I tested the arch linux aur package again on another computer today and the build worked flawlessly : )

<DataHoarder> "if you want a docker-compose..." <- Awesome! I also have an example using my own Docker images here:

Had tor setup, removed it for ease of use in the end

Oh btw sethsimmons I would recommend bumping up the outgoing connections for monerod at least

and optionally pinning some well known peers

should make block propagation quicker

Reason why tor was removed from mine was to have all images built locally, not just downloaded from dockerhub

DataHoarder: Probably for this purpose wouldn't be bad, but I generally don't recommend it.

Yeah, specifically for *this* purpose

> <@sethsimmons:monero.social> Awesome! I also have an example using my own Docker images here:

>

> github.com/sethsimmons/p2pool-docke…ob/main/examples/docker-compose.yml

Seth: if you're building over and over, might be worth giving a shot at mounting ccache's dir as a buildkit cache mount (github.com/cirocosta/monero-builds/…f5abb8c/monero/Dockerfile#L285-L286)

github.com/cirocosta/monero-builds/…c7f5abb8c/p2pool/Dockerfile#L72-L76 sech1 😅 waiting for that sweet change to easily generate static builds haha

So far I was only able to create proper static build on Alpine linux

but I confirmed it's working even on Ubuntu 16.04

niice

the problem is with getaddrinfo linking on other distros

the build itself is straight-forward - create Alpine Linux VM, apk add git make cmake libstdc++ gcc g++ automake libtool autoconf linux-headers and then build libsodium, libuv, libzmq in external/src

and then for p2pool cmae .. -DSTATIC_LINUX_BINARY=ON and so on

<utxobr[m]> "> <@sethsimmons:monero.social..." <- Cache is used on some rebuilds, but intentionally not in most others for apt upgrade purposes.

But its all Github CI so Idc about the delays due to not caching, and I don't cache only when I don't want to by design.

yeah makes a lot of sense - found quite useful when iterating constantly

Tried to generate static build of p2pool indeed, saw the way, but did not attempt on the end

Nice DNS-based seed node loading commits!

I have successfully built the p2pool and monero as per instructions in the SChernykh's github. But I am not able to get things together to mining work. Where are the binaries to run the monero daemon? Anyone known more detailed instructions to set things up?

" electricity itself costs about 0.046 kw/h, rest of the price is bulshittery" -> Tell that the people in germany ;) We re having 25 eurocent per kwh :(

@shadowfax-87 Have you followed the instructions on sech1's github page? They are pretty straight forward.

Can someone elaborate on the "Prepare enough huge pages (each of monerod/p2pool/xmrig needs them):" step in testing, I run both monerod and xmrig (soon to be all 3 with p2pool too) and haven't had to deal with manually preparing huge pages because xmrig seems to do it for me. The recommended command is "sudo sysctl vm.nr_hugepages=3072", is this good to just plug and chug in my situation?

For me it worked.

xmrpow: Yes. The sources files compiled without any errors, but where are the binaries to run monerod?

shadow: After the build they should be in the /build folder

(If it is not obvious you also need to build monerod from the linked repo, which is down a long path in build/)

abberant[m] I just skipped that step

I'm getting warnings about huge pages, idek if it matters so you're suggesting I just ignore them?

Thank you. I found it in ~/monero/build/Linux/p2pool-api-v0.17/release/bin

abberant[m]: if so, less work for me B)

abberant: what's the warning?

Hello. Have a question about RandomX. Why isn't the key change every block, say using previous block hash or some other data from previous block? RandomX doc says "the key should change every 2048 blocks (~2.8 days) and there should be a delay of 64 blocks (~2 hours) between the key block and the change of the key K. This can be achieved by changing the key when blockHeight % 2048 == 64 and selecting key block such that keyBlockHeight % 2048

== 0" but there is no detail of any security assumptions and why 2048 and a delay of 64 blocks is ideal. Maybe there was a discussion about this somewhere?

* this somewhere? (posted to monero-dev yesterday, but nobody had answer, so trying here, thx)

"... (full message at libera.ems.host/_matrix/media/r0/do…57374334cd6282c8c2c54848bbb58e66461)

on second thought maybe the last line is when it fixes itself

abberant: This was just working"sudo sysctl vm.nr_hugepages=3072" ? How much RAM do you have?

xmrpow: 16 gigs

That seems fine if it allocated in the end abberant[m] just some minor delay on share verification

would recommend having them though

and 100% for where you are mining

ok

Now is my turn with huge pages. I am getting the following message. stat /proc/sys/vm/nr_hugepages: No such file or direct

sysctl: cannot stat /proc/sys/vm/nr_hugepages: No such file or directory

do they not get automatically allocated like mine might?

> <@abberant:matrix.org> "... (full message at libera.ems.host/_matrix/media/r0/do…09cf48eccefb497a96aab3f85668c849794)

also I'm getting "JSONRPCRequest failed to connect, error ETIMEDOUT" after p2pool is left open for a little bit

my firewall drops forwards, but it shouldnt need to forward anyways since its all on the same machine right?

maybe thats blocking connection to the daemon on the same machine

shadowfax-87[m]: what kind of system are you on?

also, if you are within a jail/container you probably don’t have them unless you map them accordingly

abberant[m]: that is probably your monerod not having the right port open to connect to

mx7[m]: changing key too often means lots of time recreating the data set. Not sure why the change can't be less often though.

<DataHoarder> "abberant: that is probably..." <- any chance its the zmq port? I tried the regular rpc port since it said "jsonrpcrequest" and it didn't fix

I am using Ubuntu on Windows Subsystem for Linux.

I disabled my firewall and now instead of

"JSONRPCRequest failed to connect, error ETIMEDOUT"

I get

"JSONRPCRequest failed to parse response, result = 24"

O_O

Yes WSL will not cooperate there

as it is not Linux, at least on WSL1

it is just userland ubuntu with a windows kernel

why not compile it natively on Windows if you are on that platform?

Also do mind the bind ips might need to be 0.0.0.0, with all the confirmation and restricted flags for monerod

DataHoarder: I am getting unused with Windows. But you are right I will try natively in Windows 10.

Yeah it’s just maybe not fully right to behave it like Linux when it’s half windows in there

The nr_hugepages file for example is part of linux

the kernel