Who's filling up the mempool at the moment?

[kano]

Anyone want to analyse the mempool to work out who is bloating the transactions at the moment?

All blocks are full for a few days now, but most of that over fill (about 7000 txns as I speak) are very low fee txns.

Anyone analysed them yet to see if someone is causing it?

[o_e_l_e_o]

The mempool isn't particularly full at the moment. It did peak at 30k unconfirmed transactions a few hours ago, but it's well on the way back down now as you say, and was pretty much empty only 12 hours ago. Looking back over the last two weeks (https://jochen-hoenicke.de/queue/#1,2w), today's spike also isn't any bigger than previous ones - we've seen spikes of between 20k and 40k over the last month or so, and it always comes back down fairly quickly.

I don't think these spikes are caused by anyone in particular. Trading volume is currently the highest it's been since April of last year - it only follows that with more trading there will be more transactions.

[kano]

Yeah but the catch is that they are all VERY low fee txns.

It has indeed been dropping over the last few hours but ...
while looking at block changes, you of course see low fees when the mempool gets empty, but that was still happening with the recent block changes and showing thousands of lower fee txns not included even though the block was full.

e.g. each just after a block change, since then:
2019-02-25 22:17:45.203567 CreateNewBlock(): block weight: 3964460 txs: 1826 of 8410 fees: 0.07016318 sigops 14509
2019-02-25 22:30:55.541442 CreateNewBlock(): block weight: 3964805 txs: 2418 of 8520 fees: 0.06292218 sigops 16016
2019-02-25 22:33:52.311599 CreateNewBlock(): block weight: 3965275 txs: 1765 of 6536 fees: 0.05931998 sigops 17323
2019-02-25 22:44:11.795026 CreateNewBlock(): block weight: 3965235 txs: 1763 of 5879 fees: 0.04167562 sigops 16579
2019-02-25 23:04:23.614539 CreateNewBlock(): block weight: 3965666 txs: 2148 of 7146 fees: 0.06560297 sigops 18413
2019-02-25 23:12:57.219688 CreateNewBlock(): block weight: 3965349 txs: 1922 of 6315 fees: 0.04785037 sigops 16510

So clearly saying that the rest of the txns not used are even lower fee yet again.

[mk4]

Yeah but the catch is that they are all VERY low fee txns.

I really don't want to go full tinfoil-hat conspiracy theorist here, but along with the low fee transactions spam, bitcoin suddenly dropped in price in a very short timespan yesterday(though not that surprising), along with that, BSV had a 16% pump in the last 24 hours. Hmm. Quite fishy.

[joniboini]

Yeah but the catch is that they are all VERY low fee txns.

Deanonymization attacks perhaps? Who knows, it's really difficult to point out exactly who, but I'm pretty sure there is no one who wants to 'hijack' the block with lots of transaction with very low fees except they have some agenda behind it.

I really don't want to go full tinfoil-hat conspiracy theorist here, but along with the low fee transactions spam, bitcoin suddenly dropped in price in a very short timespan yesterday(though not that surprising), along with that, BSV had a 16% pump in the last 24 hours. Hmm. Quite fishy.

What's your guess? Are you telling that BSV supporter 'spam' Bitcoin blocks to pump their price?

[jseverson]

Yeah but the catch is that they are all VERY low fee txns.

I really don't want to go full tinfoil-hat conspiracy theorist here, but along with the low fee transactions spam, bitcoin suddenly dropped in price in a very short timespan yesterday(though not that surprising), along with that, BSV had a 16% pump in the last 24 hours. Hmm. Quite fishy.

What would the BSV camp have to gain from attacking Bitcoin though? I was under the impression that they're mainly competing with Bitcoin Cash. I know that big blocks make Bitcoin Cash more resistant to spam attacks, but even that shouldn't be a reason for them to turn their crosshairs to Bitcoin. Coincidence, maybe, or am I missing something?

Either way though, no one else seem to be talking about this. Weird.

[pooya87]

you just happened to catch this particular one, otherwise it has been happening for months. the spike size has also been the same size pushing the mempool to about 10-15 MB and drive fees up to 20 to 30 satoshi/bytes. this is one of the problems of adoption growing faster than scaling.

as for the fees of these transactions being low, the reason is because people know nowadays their low fee txs (like 1 s/b) will go through fast enough. for example during spike of today it would have taken a 1 s/b tx 12 hours to confirm in worst case scenario

[FinneysTrueVision]

Currently ~20% of transactions are coming from VeriBlock, Jeff Garzik's new altcoin, which uses proof of proof to reach consensus. There is also a new blockchain called Stacks which creates new blocks by having miners burn Bitcoin. Then there is an unknown entity which blockchain reasearcher and OXT.me dev LaurentMT has dubbed "Moby Dick" which is responsible for lots of spammy transactions.

[mk4]

What's your guess? Are you telling that BSV supporter 'spam' Bitcoin blocks to pump their price?

I'm not going to the "spam attack" conclusion just yet. My guess is just if it is indeed a spam attack, the BSV people might be the ones responsible.

What would the BSV camp have to gain from attacking Bitcoin though? I was under the impression that they're mainly competing with Bitcoin Cash. I know that big blocks make Bitcoin Cash more resistant to spam attacks, but even that shouldn't be a reason for them to turn their crosshairs to Bitcoin. Coincidence, maybe, or am I missing something?

While BCH might be their direct competitor, BTC is definitely still not out of their crosshair. I really don't know though, for the meantime, I'm just going to wait for more replies to this topic.

[pooya87]

it can not be a spam attack because the main characteristic of a spam attack is "constant" injection of transactions into the mempool and so far this does not look like a spam attack to me:



for comparison, the following is how a spam attack looks like (the picture is from Jan 2017)

pic ref: https://bitcointalk.org/index.php?topic=1776143.0

[flounderella]

Doesn't look like a spam attack to me. Maybe Venezuelans moving their hard earned money out one bolivar at a time before it becomes the next Zimbabvian dollar

[jseverson]

-snip-

Oh, thanks for that. Even if this wasn't meant to be a spam attack, it looks like it's not deliberate.

Doesn't look like a spam attack to me. Maybe Venezuelans moving their hard earned money out one bolivar at a time before it becomes the next Zimbabvian dollar

I guess it's possible, but I doubt it. I feel like Venezuelan interest in Bitcoin is overblown in general. They did detain journalists very recently (more recent than the filling up mempool) but I don't think even that's enough to trigger a spike.

[vit05]

Currently ~20% of transactions are coming from VeriBlock, Jeff Garzik's new altcoin, which uses proof of proof to reach consensus.

Some people say that could be more than 30%.


https://opreturn.org/
https://testnet.explore.veriblock.org/network-stats

“Source of the now-highest volume of OP_RETURN outputs has been identified as Veriblock “proof of proof” miners,” explained Lopp. “They are creating around 20% of all BTC transactions now."
https://news.bitcoin.com/more-than-30-of-btc-traffic-stems-from-the-veriblock-project/

[Kakmakr]

Well, the thing that worries me the most is that SegWit was supposed to eleviate most of the congestion and scaling problems and the Lightning Network should also have helped with that. If this is a spam attack or a test run for a spam attack, then we will have problems with the next Bull run, when the natural tx volume is a lot higher than now.

I think this is just a test run for a massive spam attack in the future. 

[squatter]

Yeah but the catch is that they are all VERY low fee txns.

Deanonymization attacks perhaps?

How would that work exactly?

Either way though, no one else seem to be talking about this. Weird.

I did notice a significant spike in fees earlier and was wondering what was going on, especially since the market seems so dead. When these spikes in network activity come out of left field like that, you have to wonder what's causing it. After the spam attacks of 2016/2017 I always think of the BCH camp first.

[o_e_l_e_o]

I really don't want to go full tinfoil-hat conspiracy theorist here, but along with the low fee transactions spam, bitcoin suddenly dropped in price in a very short timespan yesterday(though not that surprising), along with that, BSV had a 16% pump in the last 24 hours. Hmm. Quite fishy.

It's thought that the BSV pump is related to CoinGate (a Lithuanian based payment processor) listing BSV - https://www.prnewswire.com/news-releases/coingate-become-the-latest-payment-gateway-to-on-board-bitcoin-sv-bsv-300801257.html. Also the fact that BSV's volume was sitting at $70 million (around 0.8% of bitcoins $8.8 billion daily volume), means it's really easy to manipulate and even a small amount of buying will force the price up rapidly. I don't think it's specifically linked to what we are seeing with bitcoin at the moment.

[dothebeats]

Doesn't look like a spam attack to me. Maybe Venezuelans moving their hard earned money out one bolivar at a time before it becomes the next Zimbabvian dollar

Could be, given that bitcoin is slowly becoming overhyped in the said country, although that alone isn't an enough reason to justify what has happened on the mempool very recently.

as for the fees of these transactions being low, the reason is because people know nowadays their low fee txs (like 1 s/b) will go through fast enough. for example during spike of today it would have taken a 1 s/b tx 12 hours to confirm in worst case scenario

I'm one of those people who sends 1 s/b transactions even though my client tells me that at best, it will take 30 or so blocks before it confirms (which clearly is, most of the time, wrong). Idk, it just became a habit knowing that I'm really not expecting things to be confirmed in an instant and for me, every satoshi counts

[Atrax]

Would be interesting to know which Altcoins Garzik and his buddies support. Bittrex has a history as a vehicle for pump and dumps, and the only reason why Garzik is beating a dead horse, imho, could be that they are working on exactly such p&d-schemes.