What am I missing? Is there Blockchain/Mempool congestion that we are not seeing

[Kakmakr]

Yesterday I made 2 transactions and it was confirmed within an hour... (still slow, but still acceptable) ... today I made a tx from the same wallet and it has been almost 5 hours and it is still unconfirmed. (SegWit transaction) 

I checked the Mempool and it's not that congested : https://www.blockchain.com/charts/mempool-size  but on this site, https://bitcoinfees.earn.com/ it shows a massive spike.

I know last month Veriblock had a testnet running and 30% of the transactions was from them spamming the network.. they was also supposed to launch on the Mainnet.. (dunno what impact that will have)

Anyone?

[1714706559]

1714706559

[1714706559]

1714706559

[BitMaxz]

How much fee did you set?

Recently I made a transactions and it confirmed after a few minutes. I'm following the recommended fee from bitcoinfees.earn.com and it confirmed fast.

I think there's a little drop yesterday that is why your transactions were confirmed.


Better use this tool to analyze the current network and to check the current Mempool size
- https://jochen-hoenicke.de/queue/#0,24h

It gives me more insight and able to adjusts the period of time.

And I think global hashrate is another reason why our transaction slows down.

[fillippone]

<…>


Better use this tool to analyze the current network and to check the current Mempool size
- https://jochen-hoenicke.de/queue/#0,24h

It gives me more insight and able to adjusts the period of time.

I was looking at the same site, and coming to the same conclusion.
Jochen website is very nice, as it allows you to have a glimpse not only of the me pool size, but also on the fees paid. Fees are the price for space in block, if everyone is paying high fees, chances are that your transaction mus pay higher fees to be included in the block.
One nice way to look at this is looking at the median fee (median smooth out outliers values, that influence the mean instead).

Lock at this graph:



https://blockchair.com/bitcoin/charts/median-transaction-fee-btc?interval=1m

It graphically confirms what @BitMaxz just told you: even if the mempool is not bigger than yesterday, fees are higher.

Also, just remember that there is nothing like a “mempool”: every node see their own mempool, according to the broadcasted transaction. So the only relevant mempool is the one visible to the full node arranging block for the miner.

[ranochigo]

And I think global hashrate is another reason why our transaction slows down.

Likely just the mempool that is causing this. The block intervals, from what I observed is still slightly faster than 10 minutes so that is likely a non-factor in the delay of confirmation of your transactions.

You really shouldn't be looking at the overall size of the mempool from the start. Only the fee rates and the number of transactions within that fee range matters to you. Blockchain.com hasn't been a good source of data historically so you should really choose another data source.

[mk4]

Idunno, it seems like people moving in and out of exchanges have calmed down already. In the past few days though, especially when bitcoin first reached 11k levels again, there definitely was congestion. I had to pay like 180 sats/b to get the transaction accepted within the next few blocks.

[squatter]

I just always pony up a the recommended fee amount if its something I don't want to wait to see if it is going to confirm quickly or not.

Sometimes that's not enough. This morning, a typical mempool increased about 30-35% in size over just a few hours, with the majority of transactions paying 100-120 satoshis per byte. Directly prior to that, 20 satoshis per byte would have gotten you into the next block.

There is no fee estimation algorithm that properly accounts for this. That's what replace-by-fee is for.

[elda34b]

I still see more than 100 sats/vbyte yesterday to get your tx confirmed within the next 2 block, looks like it has calmed down rapidly this morning where 10 sats is the median fee for at least the next 3 blocks. Hopefully no more spike happens, because if that happens then I'd have to wait for Saturday or Sunday to make a tx.

[pooya87]

Yesterday I made 2 transactions and it was confirmed within an hour... (still slow, but still acceptable) ... today I made a tx from the same wallet and it has been almost 5 hours and it is still unconfirmed. (SegWit transaction) 

that doesn't tell us anything! you should look at your fee/vbyte value and then compare it with the same value of other transactions in the mempool at those times to see where you were in the priority scale.

I checked the Mempool and it's not that congested : https://www.blockchain.com/charts/mempool-size  but on this site, https://bitcoinfees.earn.com/ it shows a massive spike.

both of these sites are garbage and should never be visited under any circumstances.
blockchain.com doesn't even know what virtual byte is and their explorer is filled with mistakes and bugs.
and the other site is always reporting outrageously high fees intentionally.

[Kakmakr]

Oh, well... Thanks for the inputs. I checked again today and it was confirmed and the bitcoins are transferred... but it took almost 6 hours. I did not adjust the fee, because it was set to the same fee that I used the day before and with that fee it only took an hour to get 3 confirmations.

In future I will check first and then bump up the fees if I see other people using higher fees than normal. I think this might be a little bit too much of a learning curve for average Bitcoin users to calculate fees, luckily for me I am a bit more advanced than normal Bitcoin users.. so it is not a big deal. (Imagine the first impression from a new Bitcoin user, if they have to wait 6 hours for 1 confirmation) 

[o_e_l_e_o]

I particularly hate times like these because it takes so long for all the automatic fee algorithms on wallets and exchanges to "reset" their baseline to something sensible, which just contributes to the ongoing cycle of fees much higher than they need to be.

If we look back, there was a point a few hours ago where 20 sats/vbyte would have put you within 0.1 MB of the tip of the mempool. Less than 1 MB of transactions later, and you now needed a fee of 170 sats/vbyte to be within 0.1 MB of the tip. Every single one of those 1 MB of transactions could have paid 21 sats/vbyte and been in the next block, but instead some people were paying over 400 sats/vbyte to get in the next block.

Automatic fee algorithms on wallets, exchanges, and sites, are to blame for this. They all try to outdo each other, and just endlessly and unnecessarily push the fees higher and higher. If everyone spent 10 seconds to look at the mempool and choose a sensible fee, everyone would save a bunch of money for no reduction in confirmation time. But I know that is never going to happen.

[ranochigo]

In future I will check first and then bump up the fees if I see other people using higher fees than normal. I think this might be a little bit too much of a learning curve for average Bitcoin users to calculate fees, luckily for me I am a bit more advanced than normal Bitcoin users.. so it is not a big deal. (Imagine the first impression from a new Bitcoin user, if they have to wait 6 hours for 1 confirmation) 

Which is why most wallets uses floating fee. Using static fee doesn't guarantee the most efficient fee possible for a transaction with the given network conditions. Most wallets should be able to provide a reasonable estimate for users to follow and it helps to simplify it for most users.

In some cases, the volume might spike and result in an inaccurate fee. But those instances are not that common and are frankly quite unpredictable.

[slaman29]

People are complaining all about mempool these days but yesterday it was 27k and then max 60k I saw, and today it is back down below 20k. It's actually not THAT expensive even if you want to do next block as recommended fee was around 130 satoshis/byte, with segwit and single input that only seems expensive because of BTC price.

I paid just 10 satoshi/byte yesterday and it got confirmed a few hours ago. Wasn't urgent so I'm okay with that kind of waiting time.

[fillippone]

I guess that if you are complaining about confirmation times, you don’t have the luxury to pick up the most convenient times to fire your transactions.
Please be aware that, in addition to the all useful resources posted here, there is a very useful resource in bitcointalk.org by @tranthidung
Bitcoin transaction fees (in sats/kb). Sunday, Saturday are best to move BTC

This is a good statistical analysis for the best day to transact.

[squatter]

Automatic fee algorithms on wallets, exchanges, and sites, are to blame for this. They all try to outdo each other, and just endlessly and unnecessarily push the fees higher and higher.

There's too much emphasis on guaranteed next block confirmation. It's a cultural problem as much as a technical one.

I've withdrawn from exchanges that paid 100+ satoshis/byte when they only needed to pay 5-10, apparently just so they can never worry about congestion. It's like they're at a toll booth requiring a 25 cent payment, and they're throwing away $5 bills.

Meanwhile, the exchanges are making so much money that network fees are a write-off to them, so nothing ever changes. It's a tragedy of the commons situation. That's also why BitMEX continues to abuse the network with their daily flurry of un-optimized transactions, broadcast all at once. They have very little incentive to improve things in a timely manner.

[Artemis3]

I hate that exchanges, some pools and other online wallets don't let you choose the transfer fee. Most people can wait a day. Currently a 4 sat/B fee would take a day, it was 1.3 a few days ago. The current "next block" needs about 158 sat/B, i rarely ever need anything within the next block; but everyone uses those wallets with the fee guestimation built in, and then blame it on Bitcoin...

If you can wait a day, then use the recommended fee for 1 day confirmation. If you can wait more than a day, go with 1 sat/B.

[dothebeats]

Massive spike in fees is a result of people 'thinking' that they have to pay for more to get their transactions faster in the block, while in reality it's just an illusion that was created by past price surges since back then, the mempool is really congested and fees are staggeringly high. It has become the norm to increase fees whenever there is a massive price surge even though there isn't a problem on the mempool, and the miners love it for sure. If you can bear to wait for hours on end, you can always set your fees to a minimum and disregard the suggested fees to confirm your tx within the next block or so.

[o_e_l_e_o]

There's too much emphasis on guaranteed next block confirmation.

True, and there is also a general lack of knowledge regarding how mining and average block times work.

Meanwhile, the exchanges are making so much money that network fees are a write-off to them, so nothing ever changes.

Don't forget that 95% of the time exchanges are making profit on their network fees too. Most of them are charging anything from from 20,000 up to 100,000 sats per withdrawal. During the majority of the time when average fees are <10 sats/vbyte, and when withdrawal transactions are batched to keep their size to a minimum, these fees are a gross over-payment compared to how expensive the withdrawal actually is. I know they have consolidation transaction to do in the background, but they still make a nice profit on withdrawal charges.

Even when fees are high, as you say there is no incentive to change because their withdrawal charges are so high anyway that they still easily cover the increased fees.