supportxmr 920 MH/s, minerxmr 918 MH/s, moneroocean 320 MH/s

all at the same time

it's someone BIG

why aren't they running their own pool ...

<Donkeybrrrrrr> give them a chance,, they just learnt xmrig proxy

not like it really matters anyway

Interesting graph of nonce frequency matrix.org/_matrix/media/r0/downloa…matrix.org/HMsUUtGrHUFTfufDGhPkSJbW

damn M5M400 . 1.38 GH/s!

gingeropolous: ikr?

so when you gonna start changing the rules on us / :P

free monteros for everyone! blockreward 100!

I guess we should start DoSing that pool.

I hear there are DoS for hire services.

But seriously, it looks like PoW is really like communism. Utopia that doesn't work in practice.

stop tha A(zure)SIC

Bad channel. Guess I should say that on something with more and less technical people.

So, are you doing anything to voluntarily stop being Eve ?

Or are you going to claim "I would if I could" ?

Guess we know, since it's not new.,

redirect it to my wallet 8)

I made a variable fee years ago. Nobody gave a shit.

Anyway, time to find something else to work on.

moneromooo: what do you want me to do? kick it? so it moves to minexmr and pushed them to 80% nethash?

Yes.

Then we DoS that pool :)

last 1000 blocks has been equal split between minexmr and supportxmr

moneromooo invested in minexmr ICO confirmed

both 36%

I mean, what's the point in working on a "trust me" crpytocurrency, right ?

I'll go work on grin instead.

M5M400: it's those same 4 proxies ?

moneromooo: yeah. best to drive out the 12 people still mining on their ryzens

Ah, great, a non sequitur.

seriously though, this should die down any minute now

what if it doesn't

Till next time. Because it's happening every once in a while, does it not.

it does.

this is the largest occurence by far though

M5M400: wen 1.5% fee?

i'm with mooo, we DoS your pool and throw eggs at your house until you ban new addresses from mining (and then repeat for minexmr)

And even that is a shit solution since it's a "trust us" solution. I don't have a good one though.

can I throw eggs too?

You can aim at minexmr.

but xnbya was always nice to me!

xnbya: are you the minexmr op ? Please also up fees and/or start refusing connections when you get to like 40%.

they're both under 40% though

I mean... we can go to 110% fee

"this fee goes to 110% "

next share only accepted if paid first

we only have to nuke 40 pools before xmrpow.de is #1

Did anyone port jtgrassie's self select to other pools btw ?

I don't think I can face javascript again. Please someone do it :D

bring back minergate

xd

That'd be one excuse fewer for pool ops using other pool software.

I... think I'd rather minergate with 20%, even though they're bytecoin scammers...

moneromooo: selfselect is nice on paper but will never catch on with miners

bring back minercircle

RXboost3+

Oh, I thought the blocker was pool ops. Why wuld miners not like it, it'd be tansparent for them, no ?

))

regarding these azure spikes, i mean one thing is to make mining pools dead simple stupid to run, so these data center security audits can use their own daemon

ok, so I got some juicy data to analyze from an unknown pool op (not pointing fingers)

the thing is, they are currently not. and if you've only got n hours to use the resources, you wanna make sure it works

it looks like xmrig workers behind a proxy

3.2 kh/s average

i still have doubts my solo mining effort is working with the frigging daemon.

3.2 kh workers lol

well, xmrig pointed at the daemon

that's a lot of workers

not xmr-stak like the last time we had fat miners on nanopool

I can see very distinct pattern coming from xmrig

damn, 3.2 is like intel machines

that number is very approximate

wtf shit intel machines are that slow??

they haven't mined enough blocks to be sure

It might be anywhere between 1-6 kh/s

i guess desktop skus still have like no cache

well, consumer end intel, like my i7 something something

but these are prolly vcpus anyway

what an absolute waste

F8s on azure produce 3.2 KH/s on xmrig

and we got our suspect

it started about 8 hours ago now?

1 acccount * 1xF8s per account * 30 regions * 3.2 KH/s = 96 KH/s

900 MH/s / 96 KH/s = 9600 trial accounts

usually banned in few hours

here's 35 xmr for your troubles...

well, based on the hash increase @ moneroocean it has to definitely be xmrig

there's really no way to stop the power law, rich get richer

the more resources you have, the faster you acquire more

we would need something analagous to the dynamic fee algorithm. there, when usage goes up, fees go down.

here we want when individual hashpower goes up, rewards go down

of course, miners can just split off to separate accounts, so we can't tally their total hash power

WOuldn't something like that be extremely helpful for any sympathetic pool OP to implement? But do it on a pool-wide level?

Fees dynamically raise as % of network hashrate approaches X%, so smart miners will move, and dumb miners will lose revenue?

yeah something like that would work

shape some kind of logarithmic curve, so that once you get to e.g. 10% of network, rward goes to zero

but you can't just raise pool fee - that just further encourages these guys to use their own pools

it has to be network level, block reward calculation

But that's the whole issue, it's permissionless and anonymous, there's no way I can think of to do it on a network level

right

We can at least make it harder on them, and if they make their own pool it still helps spread hashrate among the network, increasing decentralization (unless they have 51%+ of course)

if you rely on pools to regulate, there'll always be the one pool who won't comply because greed.

yep

True, but that's not a reason to do nothing IMO

We can still do something to help encourage the spreading of hashrate

At the level of first mover/network effect advantage minexmr etc. have, it would be hard for a no-fee variant to take over

If we could get 70% of the network hashrate under a log curve of fees, that would be a big step forward to incentivize miners to move to smaller pools

And if they don't you just make more money

Seems like a win win to me

.shrug

sure. get the top 20 pools to commit to that and I'm in

anyone knows how much HR you get out of a digitalocean droplet?

If someone can write it I'll drive a campaign gladly

Across Reddit/Twitter/direct to pool ops

<M5M400 "sure. get the top 20 pools to co"> I'm not sure why you can't lead by example, honestly, but if I have to convince others first so be it.

because as I see it supportxmr is the only thing standing in the way of minexmr network dominance

not that I don't trust xnbya, but I do trust myself considerably more

So if I get them onboard you'd enable a similar fee approach?

I'd be open for it, yes.

Or I need the entire top 20?

I'll see if I can get in contact.

you will probably bite out your teeth at nanopool already

M5M400 do you know who made these graphs? imgur.com/a/0rKqdWi

is the second one logarithmic?

sech1: endor over at -pools

yes

so thick line starts at 32768, lol

that's xmrig right there, it spreads nonces between cores with step=32768

xmrigASIC confirmed!

:D

if you just make the pool fee dynamic, that means pool operators keep more of the rewards as their total hashrate goes up, if miners aren't paying attn

so it's in their interest to adopt the approach, they'll pocket more money

why does everything always have to be about money

Inge- matrix.org/_matrix/media/r0/downloa…matrix.org/HMsUUtGrHUFTfufDGhPkSJbW <-- top 8 bits are randomly selected by proxies/nicehash, and 50% of blocks have nonces < 16777216, this is why

It's about money for the miners, so the only way to change their behavior is incentives/disincentives :)

the spike at 16 (actually 15) is also because of how xmrig works

so a key question for implementation is what time window you use to estimate the network hashrate

and likewise to estimate the pool hashrate, since pools don't know miner's true hashrate

although perhaps we could add a field to the job message, for the miner to report its instant hashrate

not sure if that's actually useful, or if there's any benefit to lying

sech1: thanks

the pool can always fallback to its estimate anyway

M5M400 pool ops right now: youtu.be/hf_7xAX_fBE?t=20

:P

so... does it sound workable?

can throw self-select in at the same time

an example setup: pool op sets arbitrary fees, they apply up till 10% of network hash rate

if rate rises to 11% for >= 1 hour, pool fee becomes 2%

and doubles for every percent increase from there, to maximum of 100% at 17% of network

so miners on big pools are just throwing money at pool ops

but normal or even lower fee for self-select?

That would be pretty damn cool

do you mean self-select is completely excepted from the fee increase?

exempted

I guess that could be reasonable, since self-select doesn't feed centralization

let's just start with supporting self-select on big pools :D

with lower or no fee

why would a pool offer it for no fee?

what do they get out of it?

more healthy network?

fees can be balanced between different miners

It's like moneroocean has 0 mining fees, but gets money from withdrawal fees

ok

asymptotically did you see the fat miner at moneroocean switch to rx/wow?

or maybe it was just other miners there?

it went from 2 Mh/s wow to 70 Mh/s wow, mined a ton of blocks and then i think it went back to xmr

moneromooo/cryptonote-universal-pool 5e31901 if someone wants to run with it

hmm, 70 mh/s is not what miners there can usually do :D

ah yeah, I see blocks 235401-235431 on Wownero, almost all mined by MoneroOcean, lol

24 of 31 blocks

i wonder if you could put a throttle at the protocol level

similar to the max block size increase. i dunno what that would do to things though.

i.e., only n% is allowed in hashrate delta over n blocks

would be a bitch for pool ops to implement how to payout pplns etc

but basically, if hashrate hits some spike threshold, blockreward drops to 0

(kicks it for future, obvi)

wait, isn't this what wow is doing jwinterm ?

oh, no, that was about pool hopping

no I don't think so

wownero is going to add pseudorandom block maturity time that varies from like 1 day to 1 month

so throttling block reward via some derivative of difficulty... wouldn't necessarily modify pool centralization

but would penalize fast block producers

There was this idea of linking block reward to coinbase lock time. Lock for 6 months and get +10%. For 3 months, only +5%. etc.

so you could end up with attack miners (azure etc) mining their own chain with diminishing block rewards

Opportunistic miners would likely not want to get to a pool but not bother solo mining though.

but legit miners continue making regular speed blocks

so could lead to reorgs

i like that idea moneromooo

So what we need is an easy to use miner GUI with stupid videos showing how to clikc on a button, which connects to pools based on their hash rate.

Miners are lazy people, so let's use that laziness to our benefit. Make the easiest to use UI, and have it balance.

that doesn't really help with the guys with like 300 Mh/s on nanopool tho probably

Is monero-update easy to use ? I think I made it easy to use but I don't know what non computer literate people find easy to use.

maybe just linking block maturity time to the derivative of difficulty

diff ramping up fast, blocks mature in 2023

diff coming down, blocks mature in three blocks

rather than fiddling with reward also

That screws everyone though, no ?

Fiddling with reward screws only those who want fast unlock (ie, those who would not trust a pool to pay them in 6 months).

yes, but under the assumption that opportunistic miners would care the most

the block difficulty adjustment algorithm's purpose is already supposed to be to regulate block emission rate

throttling with emission or unlock time is just redundant work

Isnt it a little bit weird that sb can spin ub 1.4 Ghs within a day?

Do you still think that is cloud mining?

cloud mining is designed for this, no?

spin up more instances when required

I mean cloud computing

I thought it is not profitable?

it's not, if you're paying for it

if your paying for the cloud mining its probably not profitable

Even with spot instances at azure?

i dunno, u could do the calcs

I thought you can get a 90% discount

Maybe i gonna do that.

go for it

So... With spot instances you get 2 epyc vcpus for 0,0092 Euros/h (D2av4) . These should bring round about 1000khs. During one day with that hashrate and the current difficulty that should result in 0,034 Euros/day. Mining with these vcpus costs 0,22 Euros. So it s far away from being profitable.

I rly dont believe that all the big cloud companies dont see such a high load on their systems.

maybe they see it but who cares if clients pay?

it's clients' problems if their accounts get hacked

at least with aws they usually forgive costs you end up with if you get hacked

Dont think that entities with a lot of credits secure their account properly?

I think people who own aws or azure accounts are mostly tech savy.

Do you think Twitter was secured properly?

god damn Florida teen hacked it

True ;)

you can buy aws credits at 70-80% off face value with bitcoin

but even so probably not profitable even with on-demand

I think if this hashrate stays for longer then we can conclude that it is not cloud mining?

jwinterm: it's against the tos of the free credits to use it for mining (they mention cryptocurrencies explicitly), it doesn't take long to get banned

also that

but somehow these guys are regularly spinning up 10-100k instances

presumably

Dont they make additional profit by boosting hashrate while difficulty is adjusting slowly?

if their "free" credits are limited, it makes sense to ramp up when difficulty is low and stop when it adjusts

Wasnt the difficulty calculated by the last 720 blocks?

kind of

So they would have round about 24 h

less if blocks come faster which they do

looks like all moved to minexmr

minexmr 1.45 GH/s, 55.7%

crap

DoS it!

AWS didn't care for the longest time. Used to be great because you could buy AWS codes for 10% or less of face value, w/ spot instances, you'd double-quadruple your cashe without too much of a problem.

Because by the time you were paying < 1% of face value, it was more than profitable to mine.

if it was possible would there be anything against getting rid of user pools and only having network pool(s)? distributed and powered by monerod or another daemon if needed.

kinghat[m]: how do you prove it's just the daemon mining and not anything else like a bunch of xmrigs connecting to it?

Snipa: But you cant buy that now right?

well there would only be one network pool so you wouldnt have to worry about a user pool getting 51%.

Rumor: i.imgur.com/ILTBkcP.png

xmrpow - Of course you can?

You just have to be ready to roll through AWS accounts and do some heavier lifting.

Snipa: Why should aws sell credits under value?

They don't sell them, they hand them out for /free/

Communism!

You mean to startups etc?

Go hit the massive bitcoin sales forum, and check it out. People collect them in bulk from events, then resell them at a fraction of face value.

Because they got them for free.

Used to be able to buy 10-15k/month per AWS account w/o a problem.

But yeah, people used to make fake AWS accts too, get the 150k credit and sell it for 10-15k.

there you go

Way back on CN, wolf0 had a chunky private miner for AWS that had even more profitability on AWS instances.

M5M400 same address just switched pools? He didn't like you?

maybe I leaked too much info

If that guy wanted to hide, he wouldn't mine on a pool, lol

82oiMVmcV8W7yhWeK2hiDZLVNxwHcugNafCSzk9Zbs3p645n7gbHqf4TKHXrMTHXYPQffgZ9TUebKTr5ZfRN5arV4Vjtvko

oops

:P

lol

Lets adjust the payout threshold to 0.01 ;)

Does the free payout threshold at minexmr autoadjust?

I set it to 50

because why not?

Why not 0.5

more transactions

;)

Let's hope he won't mine to 50 XMR :D

sech: You rly dont like the 0.5 ? ;)

minexmr should really use password for changing these things

yes

now everyone will mess around there

sech: True.

Im curious if these guys are monitoring this channel ? ;)

probably

I'd almost be more surprised if they weren't.

Sadly the payout now button has a cooldown :-D

lol

you devils

I dropped them a line at their workerid:-D

what does it say? "come to my pool, bitch?"

Not that rude :P

pussy

:P

oh, sorry. this is not -pools ;)

I go stand in a corner for using that language

I'd expect them to change to a differnet address if you anoy them enough

and idk about passwords, didn't you drop them M5M400

I did. smooth sailing support wise since

Yup, because people couldn't remember them.

Or figure out how to set them.

Or really, understand how to mine effectively.

xnbya: hurr durr raise fees ban ban durr

they'll stop soon i suspect

he has 5-7% expired shares

very inefficient proxy

DO droplets probably maxxed out

wonder why he switched

should've used xmrig-proxy, lol

our luck has been consistently berrer today :)

Hashrate already dropping.

*better

max was 950

the one thing I don’t get, why do these cloud providers don’t fix their security

these spikes have been going on for a while

yeap. they don't seem to care

Maybe he is switching address.

or runs out of steam

long overdue

Also, there's a huge difference between "Cloud providers need to fix their security" and "Developers need to stop leaving private keys in repositories"

lol

There's tools that literally just walk github looking for AWS and MS API keys.

Though I think github finally put in some protections against uploading them.

never underestimate the stupidity of "developers"

Particularly "senior" developers. :D

especially in docker generation

xnbya: lol I rat out his address and he jumps back to us?

almost certainly unaware of it

well, he didn't get a payout on minexmr because I set threshold to 50 XMR

yeap

and decided to switch back, lol

WHY DIDNT HE GO TO XMRPOW.DE?????11

Lulz.

he doesn't know about dashboard?

Yes ;)

xmrpow: I feel for you <3

Im trying it again:P

bur... it would instantly make you a megapool

you hate megapools

I do :)

But it is cheap marketing ;)

but you could use a block. fair enough

M5M400: Maybe he is going to bite this time :P

fingers crossed, I guess?

just remember it's temporary. don't make the same mistake as respectXMR did and already spec your lambos color

"WhyNotXmrpow" would be better :D (twitter hack message)

sech1: Im going to improve my marketing ;)

M5M400: With 0% fee a lambo will be difficult :)

Snipa: black. black nismos matter

xmrpow: true.

M5M400: Block would be totally fine, because I still have ton of shares in that round.

Somebody has stolen my idea :)

what doeas it say?

MoneroOceanFTW

lol

:-D

M5M400: Are you instantly removing not working rigs?

no. there's a grace period

how long?

no idea

M5M400: you should just start relaying shares to smaller pools. transparently

ugh. still no solo blocks

jtgrassie's pool is mocking me with its 296 blocks

2.8GH/s

Crazy...

Hey... reporter here... Anyone has 5mins to talk to me about a botnet currently impacting the hashrate?

What info are you looking for? What's the core of the story? How do you know it's a botnet and not just someone abusing free Azure credits?

That's what I'm here to do. Gather leads, info from some monero pool operators, and go from there. Just following a tip for now.

Best contacts would probably be sech1

or M5M400

But depends on what you're looking for.

hyc: I would rly like getting more hashrate, but does it rly solve the problem?

Relaying to smaller pools is worse than bumping fees or rejecting, since there's still the possibility to commandeer the hash rate without notice.

hyc: M5M400 would decide as a central entity about small pools getting hashrate or not.

moneromoo: I agree, variable fee would be a better option.

campuscodi40: It’s not a botnet, it’s someone abusing cloud providers, these hashrate spikes usually die down quickly

campuscodi40: as I noted in my tweet at least a portion of this if not the whole thing is on Azure. my original estimate of servers is too low

Even if it's not a "botnet" in the classic meaning of the word (as relying on malware), it's still of interest, just because of the mass-abuse of cloud infrastructure

yep

but I don’t think the same attack vector was used this time

My feeling is Monero is still "small" enough that reputations and ethos matters significantly compared to pure economics, and with enough social pressure we can enforce the variable fee.

Might not be possible forever, but I think we're still there

"enforce" to be interpreted in the soft sense of course.