What happened here? jochen-hoenicke.de

[AB de Royse777]

https://archive.vn/wip/kXwfI

As you see there is a break/drop of everything and it's very unusual size for every sat range too. Is this something related to actual pool or the site owner had a break of connection or maybe something gone off in his hosting/domain?

Nothing very serious but just curious to know the reason.

Cheers,

[1715170762]

1715170762

[1715170762]

1715170762

[1715170762]

1715170762

[eaLiTy]

As you see there is a break/drop of everything and it's very unusual size for every sat range too. Is this something related to actual pool or the site owner had a break of connection or maybe something gone off in his hosting/domain?

Nothing very serious but just curious to know the reason.

I was expecting that my transaction would confirm when the graph dropped down but it was disappointing as it is stuck for the past few days  and i do not think that it was a break in connection, yesterday or day before i saw a thread here talking about a billion dollars worth of coins getting confirmed for a few dollars when the entire mempool was full and the transaction charges were well above 700 SAT/B.

Here is the thread.
Bitcoin Whale transfers $1 billion in bitcoin and pays $3.54 in fees?

[NeuroticFish]

Nothing very serious but just curious to know the reason.

I've just noticed that. It's also present on btc.com chart and it made me curious too.

a thread here talking about a billion dollars worth of coins getting confirmed for a few dollars when the entire mempool was full and the transaction charges were well above 700 SAT/B.

That transaction has happened last Sunday. This "glitch" is from the last 24h (last night). So there's no relationship between them.
If you read through that tread you'll see that the fees were smaller last Sunday and that transaction simply paid the normal fee.

[pooya87]

the owner is also running a bitcoin full node on a computer that can crash, be shut down/restarted,... for many reasons. and when you do that your node has to come back and report back to the front end (the website) about its mempool and all the fee rates which is what must have happened here.

I've just noticed that. It's also present on btc.com chart and it made me curious too.

i wouldn't be surprised if btc.com were taking its data from jochen-hoenicke.de!

[NeuroticFish]

the owner is also running a bitcoin full node on a computer that can crash, be shut down/restarted,... for many reasons. and when you do that your node has to come back and report back to the front end (the website) about its mempool and all the fee rates which is what must have happened here.

That's what I thought until I've seen it in the second place. Now I wait for more info. Did more research, it has to be a crash.

i wouldn't be surprised if btc.com were taking its data from jochen-hoenicke.de!

LoL !!!!  
I would expect better than than that. I mean, they have their own pool, why would they use 3rd party data?
Also btc.com chart may be older that J-H one (at least I've heard of btc.com page way earlier than about the J-H page, but I am aware that this may not mean anything).

On the other hand.. other 2 charts don't seem to have the glitch:
https://mempool.observer/#historical-mempool
https://www.blockchain.com/charts/mempool-size


My current assumption is that certain software has crashed. If it was the same computer or not, I cannot tell.

[AB de Royse777]

I was expecting that my transaction would confirm when the graph dropped down but it was disappointing as it is stuck for the past few days  and i do not think that it was a break in connection, yesterday or day before i saw a thread here talking about a billion dollars worth of coins getting confirmed for a few dollars when the entire mempool was full and the transaction charges were well above 700 SAT/B.

Here is the thread.
Bitcoin Whale transfers $1 billion in bitcoin and pays $3.54 in fees?

Not everyone is making $1 Billion dollar transaction :-P

Earlier I had a tx with 150 sat/Byte got confirmed and it was hanging up for last two days I guess. But this is the reality we have to accept I guess.

On the other hand.. other 2 charts don't seem to have the glitch:
https://mempool.observer/#historical-mempool
https://www.blockchain.com/charts/mempool-size

I thought pooya87 had a good explanation.

My current assumption is that certain software has crashed. If it was the same computer or not, I cannot tell.

You mean more than one software? If it happens for only one software then this must be a massive one and I wonder which one was that.

[NeuroticFish]

I thought pooya87 had a good explanation.

It was not bad at all, but I think that the chances btc.com uses jochen-hoenicke.de for data are not so big (of course, I may be wrong).
Hence I think that both were using the same software somewhere and that one has crashed for both.

You mean more than one software? If it happens for only one software then this must be a massive one and I wonder which one was that.

I am not sure what you mean. You know that same version of same program can be copied, installed and ran on different machines, right?
I guess it could/would be something in between the bitcoin core and a database where the website gets its data from. But I am just assuming.

[tranthidung]

yesterday or day before i saw a thread here talking about a billion dollars worth of coins getting confirmed for a few dollars when the entire mempool was full and the transaction charges were well above 700 SAT/B.

Here is the thread.
Bitcoin Whale transfers $1 billion in bitcoin and pays $3.54 in fees?

That transaction was made on 2020-10-26 15:19 UTC. It was about exactly 4 days ago (as of writing, it is 18:48 UTC 30 Oct 2020).

• Replace-by-fee was not chosen by the sender so you can get vague idea that sender did know what (s)he was doing.
• Fee rate: 69 satoshi/byte.

With the 1 week window, you can check more details: https://jochen-hoenicke.de/queue/#1,1w

That transaction would be an instant one with first confirmation in next 1 or 2 blocks. The screenshot (at 15:11 UTC) is not totally correct because around 15:18 UTC, there was less than 1 MB in mempool from the tip to 70 satoshi/ byte fee rate. I don't know when the sender sign that transaction but it seems to be a very quickly confirmed transaction and it was made when the total size of mempool is about 10 MB. It was done before the transaction fee storm.

[ABCbits]

There's similar discussion at https://bitcointalk.org/index.php?topic=5284844.0 few days ago.

I am not sure what you mean. You know that same version of same program can be copied, installed and ran on different machines, right?
I guess it could/would be something in between the bitcoin core and a database where the website gets its data from. But I am just assuming.

Do you mean source code of jochen-hoenicke.de (https://github.com/jhoenicke/mempool) and there are bug which makes both website stats looks weird.

[logfiles]

LoL !!!!  
I would expect better than than that. I mean, they have their own pool, why would they use 3rd party data?
Also btc.com chart may be older that J-H one (at least I've heard of btc.com page way earlier than about the J-H page, but I am aware that this may not mean anything).

On the other hand.. other 2 charts don't seem to have the glitch:
https://mempool.observer/#historical-mempool
https://www.blockchain.com/charts/mempool-size

Mempool.space's graph have doesn't have such an error as well https://mempool.space/graphs#24h

It's such a shame for Btc.com thou. And they even went ahead and put a BTC.COM label on their chart 

[NeuroticFish]

Do you mean source code of jochen-hoenicke.de (https://github.com/jhoenicke/mempool) and there are bug which makes both website stats looks weird.

Indeed, it may be that or something heavily based on that.
Good find, I didn't pay attention to see that the source code for J-H site is on Git.

[PrimeNumber7]

As you see there is a break/drop of everything and it's very unusual size for every sat range too. Is this something related to actual pool or the site owner had a break of connection or maybe something gone off in his hosting/domain?

Looking at the chart itself (without looking at block data), I can tell you this is a glitch in the software the operator of jochen-hoenicke.de is running and is not a reflection of the state of the network. A node's mempool size will decrease each time a block is found. The mempool will decrease more quickly when two blocks are found in quick succession. After several blocks are found in quick succession, the mempool size will gradually increase.

In order for the line down on the chart to be real, at least 20 blocks would have had to have been found in the time one datapoint is collected on the graph. The line on the chart going back up is simply not how the network behaves.

I reviewed found blocks from Oct 29, and in fact there were not blocks found that resulted in all those pending transactions being confirmed at nearly once.