<s​tonekell:matrix.org> Hello all... I am new to mining crypto... my computer seems to be struggling with CPU power, my Monero wallet / daemon are now properly synced, however Gupax is not working though... dunno whats to be done here... are there modifications via command lines that will make best use of my setup?

how much RAM do you have

and what is your CPU

<s​tonekell:matrix.org> System 76 linux:

<s​tonekell:matrix.org> matrix.monero.social/_matrix/media/…matrix.org/BmRWElgeohxHosgFPgslspkv

when will they fix matrix attachments for the IRC bot

<s​tonekell:matrix.org> ?

is it known how to modify the link to read the content?

<s​tonekell:matrix.org> can you see my attached image?

␌tonekell: matrix␌long msgs and similar arrive here as link, but they are all 404

last one being this --> matrix.monero.social/_matrix/media/…matrix.org/BmRWElgeohxHosgFPgslspkv

<a​lexandre:uii.pt> (image has i7-10710U with 64GB)

oh thank you for describing the image

but that.. shouldn't stop Gupax

is p2pool and xmrig on?

<s​tonekell:matrix.org> yes m-relay they are on. I can copy paste relevant gupax message lines... if'n you would like...

can you check both xmrig's logs and p2pool's logs?

oh oh!

maybe p2pool is busy syncing up the blockchain!

<s​tonekell:matrix.org> matrix.monero.social/_matrix/media/…matrix.org/YjkDFwPLBslDmVdNGTysIWYc

.... another description please?

<s​tonekell:matrix.org> it won't let me copy / paste info from gupax interface...

<b​asses:matrix.org> p2pool couldn't start ZMQ reader

<a​lexandre:uii.pt> failed to connect to 127.0.0.1:18083, did you setup your local node properly?

<s​tonekell:matrix.org> here is gupax XMRig page:

<s​tonekell:matrix.org> matrix.monero.social/_matrix/media/…matrix.org/JuXtDHKsGsaNNJrpffjsYywy

if you're not using a local node, then use a remote one

<b​asses:matrix.org> Monero address shouldn't be empty?

address in p2pool shouldn't be empty, yeah

but is your localnode working?

<s​tonekell:matrix.org> my local node is my monero wallet GUI, set to solo mining and all synced up with chain itself, Gupax has proper primary asddress. damn my computer is very slow...

uhhhh

that's not how it works

<a​lexandre:uii.pt> if you are gonna use p2pool/gupax then you shouldn't be running solo mining, you have to enable zmq

configure monero wallet GUI to use a local node

don't use solo mining within wallet GUI

<s​tonekell:matrix.org> like i said... brand new to all of this...

<a​lexandre:uii.pt> github.com/hinto-janai/gupax#running-a-local-monero-node read the locs ;)

<a​lexandre:uii.pt> *docs

<s​tonekell:matrix.org> ok,

also btw

how much space do you have

in your computer

and are you running on a HDD or SSD?

<o​frnxmr:xmr.mx> Did u do what i said yesterday

<s​tonekell:matrix.org> i don't always understand what is being said in these guides. not crypto savvy. i'll go look though...

depending on your drive, it might be unfeasible to run a local node

or not

<s​tonekell:matrix.org> that's what i'm wondering too...

<o​frnxmr:xmr.mx> I told you what 2 do yesterday

if you have a few gigabytes, you can run a pruned blockchain, but if you have a HDD, it's probably best you don't runa local node

get away with a pruned blockchain*

<o​frnxmr:xmr.mx> cindy, my understanding is that stone is already running a node via monero gui

<o​frnxmr:xmr.mx> And is already synced

<s​tonekell:matrix.org> how do i enable ZMQ?

<s​tonekell:matrix.org> correct, ofrnAI

<a​lexandre:uii.pt> read the doc I sent... it's quite straightforward, even has a picture

<o​frnxmr:xmr.mx> That doc wont help with gui

<o​frnxmr:xmr.mx> I sent instruction yesterday

<o​frnxmr:xmr.mx> let me scroll and find it

<a​lexandre:uii.pt> no? it shows how to enable it on the gui 🤔 then it should be fixed if it doesn't work

<e​longated:matrix.org> Enter --zmq-pub=tcp://127.0.0.1:18083 into Daemon startup flags

<o​frnxmr:xmr.mx> thanks elongated

<s​tonekell:matrix.org> ok, flag entered... my wallet is connected, no longer mining since i swithed it P2Pool

<o​frnxmr:xmr.mx> Dont do anything on the mining tab in gui

<s​tonekell:matrix.org> should I start up mining again?

<o​frnxmr:xmr.mx> no, not with gui

<o​frnxmr:xmr.mx> Start mining using gupax

<o​frnxmr:xmr.mx> After entering flag, you should stop -> start the daemon (in gui)

<s​tonekell:matrix.org> ok... is that going to take hours to sync once I restart, like it did yesterday?

<s​tonekell:matrix.org> Thank you all for your help! I'm taking notes.

<o​frnxmr:xmr.mx> No

<o​frnxmr:xmr.mx> A few seconds

<s​tonekell:matrix.org> ok, did it. seems to just restart automatically. sticll connected. now goto gupax?

<s​tonekell:matrix.org> ok, here's scrn shot of gupax after gui daemon restart...

<s​tonekell:matrix.org> matrix.monero.social/_matrix/media/…matrix.org/osVzmaeISMXqRjMmaczFlzsB

<s​tonekell:matrix.org> Everyone's gone quiet. hope I didn't upset anyone. not sure where else to turn for advice... I'll go read guides / research. don't mean to take up all your time.

i wish i could help

but i can't look at matrix attachments

<s​tonekell:matrix.org> hmm, gupax won't let me copy n paste info. maybe I should set wallet to remote node? unsure. i'll go research.

<s​tonekell:matrix.org> or, just mine with wallet, not gupax.

<o​frnxmr:monero.social> No, just that time of day

<o​frnxmr:monero.social> What are the rest of your p2pool settings

<o​frnxmr:monero.social> (In gupax)

i wonder what happens if monero starts associating your private spend key with mining

like wownero

to completely kill off pools

<o​frnxmr:xmr.mx> The same thing that happens in wownero. botnets become solo-miners

but honestly i doubt qubic will take over shit

you think the amount of people who use monero, and they'd let some memey cryptobro try to take over?

he should probably quit while he still can, because there are people he's not gonna want to piss off

<m​odul8:matrix.org> It would be good if gupax can handle this or tell you how to start monerod for p2pool mining.

Is there any way to get a GPU to mine monero?

<r​avfx:xmr.mx> Yeah, xmrig support opencl and cuda

<r​avfx:xmr.mx> But you are going to waste a lot of power for not much

<r​avfx:xmr.mx> my Radeon RX 9070XT is equivalent to what? 2 CPU cores while taking 220W

<r​avfx:xmr.mx> and I can mine with 32 thread with ~100W

<r​avfx:xmr.mx> soooo

<r​avfx:xmr.mx> Use the GPU if you want but know that it's extremely inneficient to do so (by design)

i don't think you should start monerod, p2pool and xmrig all on the same computer

monerod has its own RandomX verification code, p2pool too, and xmrig also needs memory for RandomX

unless you have enough RAM for all that, split between servers

(i think p2pool can offload RandomX verification to the monero node, but still)

<k​ufecure:xmr.se> hi there, which mobile monero wallet is reproducible?

<f​ijxu:nadeko.net> what do you mean with reproducible?

<f​ijxu:nadeko.net> like NixOS or what

<f​ijxu:nadeko.net> lol

<f​ijxu:nadeko.net> MONERO WALLET, PLAY MUSIC

<f​ijxu:nadeko.net> sory

<b​asses:matrix.org> none

<1​7lifers:matrix.org> they can't do that

<1​7lifers:matrix.org> xd

<b​asses:matrix.org> quite sad, worrying that no one could do it

<1​7lifers:matrix.org> lmao

<e​ddie:oblak.be> is that not it a piece of cake? 😬

<f​ijxu:nadeko.net> PIECE OF CAKE?

<1​7lifers:matrix.org> i want some

<1​7lifers:matrix.org> mmm

So mining as an unprived user, 'monero', that wallet can not be also accessed by my regular user? As the monero user has his own dir with the wallet files, they cannot be accessed locally much less over the LAN?

Ah, don't worry about me anymore. I've got it.

i had a question related to this

is it really recommended to run XMRig as root?

Certainly not from a security standpoint. But it is necessary to set CPU MSRs and a couple other things.

XMRig has seperate scripts for setting the CPU MSRs and hugepages manually

can i run those as root, and then XMRig as non-root?

(if they still work)

Seems plausible. The main thing is turning -off- CPU Prefetch (thru MSR). Personally I did this in BIOS under Advanced.

Prefetch plays havoc with the XRandom algo.

Can also set hugepages to 1GB in /etc/sysctl.d/whateveryourfilenameis.

1GB really depends on how much RAM you have

Ya I have 64GB and that's what it wants.

<o​frnxmr:xmr.mx> "<C​indy> is it really recommended to run XMRig as root?" < yes

<o​frnxmr:xmr.mx> "<C​indy> can i run those as root, and then XMRig as non-root?" < yes

Citation?

<o​frnxmr:xmr.mx> I know you're not talking to me (?)

Why yes, I am. Recommending ppl set to run as root.

<o​frnxmr:xmr.mx> Put me back on your ignore list 💢💢💢

i like to reduce the attack surface though

XMRig does not need root after the initial CPU setup

Good move.

i'd like it if XMRig had an option to run itself as unprivileged after setting up the CPU registers automatically

like how browsers usually run child processes

Depends on how you run it. If in Gupaxx it slams you into sudo, which I never use. If on CLI you're free to decide.

i know it's easier and convenient to give XMRig unrestricted root access lol

# su monero && xmrig ...etc.

And o​frnxmr: you did it to yerself, kid.

Huh, I wonder whether /dev/hugepages must be writable by the xmrig process?

<o​frnxmr:xmr.mx> I did what? Un-ignored myself?

quantum`: hugepages can be allocated without root

using madvise

but you need to set the kernel parmeters to enable them

I'm researching how to set the size in sysctl.d. I can set the number with vm.nr_hugepages = 3072.

Another question though is does xmrig need rights to /dev/hugepages?

that would need a review of the source code, but i dobut it

since it doesn't need to share the memory... since it's one process

I am finding only a kernel param way of setting hugepage size. I guess this makes sense.

If I give the VM 48GB I should be safe allocating 40 hugepages of 1GB.

Idk how to tell whether you're memory-constrained while mining.

maybe you can tell by the fact if it uses swap

or if it fails to allocate, you need to disable memory overcommitment

for this though

I'll be darned, it actually booted...

booted without root?

No I made some mods to the kernel commandline tweaking hugepages

ah

Checking...

It twerked! # cat /proc/meminfo | grep Huge

Now, has hashing improved?

xmrig has a shell script that sets CPU registers

you should run that as root too, as see if xmrig works as-is without sudo

I've allocated 48GB to the KVM VM and set: hugepagesz=1G hugepages=25 default_hugepagesz=1G

ooh

tbh i thought 1GB hugepages only provided like

tiny benefits to the hashrate

Oops the number should have been 40, but my sysctl.d file fixed that.

I can take out that sysctl entry since I've fixed the kernel line.

If you're memory-constrained I've gotta think it makes a big diff.

I discovered I'd allocated only 16GB with 55 CPUs.

syncing...

Ok, HUGE PAGES supported, but 1GB PAGES disabled, running as unprived user.

Ok, HUGE PAGES supported, but 1GB PAGES disabled, running as unprived user.

No change in hashing speed. It must be only using 8GB of RAM.

Trying as root...

<o​frnxmr:xmr.mx> Diary, when?

Huh, 1GB HUGEPAGES still not supported. (as root)

Check if your CPU supports them

sech1: is it a specific CPU extension?

cat /proc/cpuinfo | grep 1gb

Yes, cpu reports it

Don't remember the exact name, but it has 1gb in it

pdpe1gb

You should try to enable them in grun, and allocate 3 pages

grub

i mean, since quantum` is running in KVM

if they switch off KVM, and emulate pdpe1gb, it could work

(but worse hashrate)

Need to set "1gb-pages": true, in xmrig json file.

And maybe huge-pages-jit.

See any problems? zerobin.net/?902521b461b715f0#QFidg…43Wc2uXYZURMzZBRne+rvS7qK0XCvoQAq4=

can you check /proc/cpuinfo

For example: zerobin.net/?4a6f9a0bdea8716a#wtUTq…/sNcMb5FSdjvdUfiLsSWlHD1JxQtD50xoY=

I'v pinned CPUs so they're not thrashing cache.

... and turned off Prefetch in the host's BIOS.

CPU is passthrough to the guest.

Fixin' to try my shiny new .json...

okay i see pdpe1gb in the flags

Oops, crash on "coin"...

Apparently monero is invalid.

HA, just needed double-quotes.

Rockin' HUGEPAAGES now.

GODDAM, TRIPLED my hashrate! And still using only half the CPUs for some reason.

how much is your hashrate now :P

54 CPUs, 10,111 H/s.

... nd using only half my CPU capacity for some reason.

that is like 187 H/s per CPU

which is weird

is your CPU designed like a GPU?

like more cores but each one weaker?

No, standard server. (EPYC). top shows CPU 2,907%

Oh, could be disk-bound. Rotating drive.

Wow, what a difference though.

it shouldn't be disk bound

unless you have swap

in fact, mining doesn't use the disk at all

Turned off swap, intending to move to RAM but never got to it.

well that should be the issue

You're right, iotop shows the disk hardly used.

... and I'm mining as non-prived user monero.

So now I have some new mysterious constraint.

The CPU can run full-bore standing on its head, with the large eatsink I have.

also btw

did you set the MSR registers?

No, haven't bothered as I thought that is all about prefetch.

ah

well fair

Rebooting to get swap...

Oops, no mo' swap part.

Gotta research frickin' swap in RAM...

swap.. in RAM

Yup, dun did it befo', gotta re-learn now.

"Note that there are kernel modules Zram and Zswap which allows compressing the swap inside RAM. Zram does not need any physical swap device."

Well that was easy. # aptitude install zram-tools systemd-zram-generator

zram is gonna slow down your hashrate :P

since it needs CPU to decompress and compres

IF it were using swap, which it shouldn't.

zram still needs CPU to decompress the compressed RAM

zram has no appreciable effect on hashrate.

Only when it's used tho, right?

yeah, only when the compressed memory is used

Oh sure, it says, "FAILED TO APPLY MSR MOD, HASHRATE WILL BE LOW", but it doesn't really mean it.

free -h shows zero swap used.

did you benchmark hashrate between zram and no zram?

Essentially. I remember what it was before, and now.

It's a hair lower: 10,613

Still not using half my CPU.

Ok stopped /dev/zram0. (systemctl stop /dev/zram0)

Uhnh, 10,447 now

Something is mightily constraining my bandwidth. Maybe htop will advise...

I have a question about shares diff, I have xmrig-proxy connected to p2pool, p2pool gives to high diff, I set it manually for each miner, so shares happen approx. once per ~30 secs, I have no idea what I'm doing but I read somewhere it should be like this (ideally 1 share per 30 secs)?

should have asked in monero-mining I guess

Found it.

ruidx p2pool auto-adjusts diff, and you don't need xmrig-proxy to connect to p2pool

but yes, one share per 30 seconds is fine

nice, I just have a little UI for xmrig-proxy using http api, like to watch tables with stats :)

<1​7lifers:matrix.org> scam alert - fake tradeogre website ibb.co/xq94WCjS

homer

simpson

I cannot find it in english, but this was the scene --> inv.nadeko.net/watch?v=vokA5OpKaQA

<b​asses:matrix.org> blog.accrescent.app/posts/the-future-of-accrescent

<b​asses:matrix.org> Monero was the highest donation

<1​7lifers:matrix.org> as it should be.

<b​asses:matrix.org> Monero has always been #1 biggest supporter to FOSS projects

<1​7lifers:matrix.org> rip to those who paid with eth and btc

<1​7lifers:matrix.org> theyre logged and will be hunted down at some point by any such entity

<e​ddie:oblak.be> Anyone here know this guy? youtube.com/@ShadowofAtlas He references XMR sometimes. I am not sure if all he says makes sense, but it's entertaining content anyway.