WARNING: abnormally high number of blocks generated

[chek2fire]

I see this message today to my nodes

    "errors" : "WARNING: abnormally high number of blocks generated, 48 blocks received in the last 4 hours (24 expected)" 

and after that i see this message from luke-jr in reddit

https://www.reddit.com/r/Bitcoin/comments/3ydwg2/warning_abnormally_high_number_of_blocks/#oo

Discussion thread for this new warning.

What this means:

48 blocks were found within the last 4 hours. The average is "supposed" to be 1 block every 10 minutes, or 24 blocks over a 4 hour window. Normally, however, blocks are found at random intervals, and quite often faster than every 10 minutes due to miners continually upgrading or expanding their hardware. In this case, the average has reached as low as 5 minutes per block, which triggers the warning.
If the network hashrate was not increasing, this event should occur only once every 50 years. To happen on average, persistently, the network would need to double its hashrate within 1 week, and even then the warning would only last part of that 1 week. So this is a pretty strange thing to happen when Bitcoin is only 6 years old - but not impossible either.

Why is this a warning?

It's possible that a new mining chip has just been put online that can hash much faster than the rest of the network, and that miner is now near-doubling the network hashrate or worse. They could have over 51%, and might be performing an attack we can't know about yet. So you may wish to wait for more blocks than usual before considering high-value transactions confirmed, but unless this short block average continues on for another few hours, this risk seems unlikely IMO.

Has the blockchain forked?

No, this warning does not indicate that.

Will the warning go away on its own?

Bitcoin Core will continue re-issuing the warning every day until the condition (>=2x more blocks) ceases. When it stops issuing the warning, however, the message will remain in the status bar (or RPC "errors") until the node is restarted.

Is this related to some block explorer website showing the same blocks twice?

No, as far as I can tell that is an unrelated website bug.

what is this warning meaning for bitcoin? Is there something unusual happen?

[Meuh6879]

not usual ... but not surprised result, too.
noticable is the right term, i think.  

13 blocks in 1 hour is high, yes but not a real problem when you have 1 block in 1h30 sometime ...  

and so, the difficulty is adjust more rapidly when blocks are generate quickly.

[chek2fire]

i have reload my node and everything is back to normal. Is the first time that this happens?

[NorrisK]

i have reload my node and everything is back to normal. Is the first time that this happens?

Most likely it is not. Sometimes because of chance a higher number of blocks can be found in a short period of time. It may also take a lot longer.

As there are not mystery miners, it is not because of a huge increase in mining capacity or something. Probably just a random event that can occur every now and then.

[rienelber]

not usual ... but not surprised result, too.
noticable is the right term, i think.  

13 blocks in 1 hour is high, yes but not a real problem when you have 1 block in 1h30 sometime ...  

and so, the difficulty is adjust more rapidly when blocks are generate quickly.
[/quote]

Exactly, it's statistically "normal" that this happens.

[ATguy]

If the network hashrate was not increasing, this event should occur only once every 50 years.

Why is this a warning?

It's possible that a new mining chip has just been put online that can hash much faster than the rest of the network, and that miner is now near-doubling the network hashrate or worse.

what is this warning meaning for bitcoin? Is there something unusual happen?

We know the hashrate is increasing, maybe up to 10%, so this mean this event can happen more often than once every 50 years just by chance. But even once in 25 years is lucky event indeed.

Normally the farms add signifficant hashing power over longer window than 4 hours, it seems pretty impossible someone double the network hashrate by installing all the ten thousands of units and then just power it all up 

[saturn643]

It looks like the warning is gone now.

[VeritasSapere]

If the network hashrate was not increasing, this event should occur only once every 50 years.

Why is this a warning?

It's possible that a new mining chip has just been put online that can hash much faster than the rest of the network, and that miner is now near-doubling the network hashrate or worse.

what is this warning meaning for bitcoin? Is there something unusual happen?

We know the hashrate is increasing, maybe up to 10%, so this mean this event can happen more often than once every 50 years just by chance. But even once in 25 years is lucky event indeed.

Normally the farms add signifficant hashing power over longer window than 4 hours, it seems pretty impossible someone double the network hashrate by installing all the ten thousands of units and then just power it all up  

Yes there is nothing to worry about this is just variance, to be expected considering that the discovery of blocks is in part random. This is not because of some miner turning on lots of hashpower, we would have noticed that if that was the case. One of the advantages of a transparent blockchain after all.

https://blockchain.info/en/pools
https://bitcoinwisdom.com/bitcoin/difficulty

[makingwin1]

lets hope it will not impact the price to go down because of that

[VeritasSapere]

lets hope it will not impact the price to go down because of that

If anything this statistical anomaly is a good thing, since people will get their transactions confirmed faster while also helping to clear out the backlog in the mempool.

[Amph]

this is like saying, "we had an average of 1 block every 5 minutes for the past 4 hours", double of the luck, would have been an issue in case it was every 1-2 minutes maybe...

as i see it it can happen, like it happened in the past that for two hours there were no blocks

p.s. yeah my bad 4 hours

[amacar2]

this is like saying, "we had an average of 1 block every 5 minutes for the past 4 day", double of the luck, would have been an issue in case it was every 1-2 minutes maybe...

as i see it it can happen, like it happened in the past that for two hours there were no blocks

I have also seen lots of delay in past finding blocks and also i have waited for more than one and half hour to get my transaction confirmed. I think this recent fluctuation is to balance those lag in past. It is good to know that network is self balancing.

[franky1]

this is like saying, "we had an average of 1 block every 5 minutes for the past 4 day", double of the luck, would have been an issue in case it was every 1-2 minutes maybe...

as i see it it can happen, like it happened in the past that for two hours there were no blocks

4 hours not days..

i think luke_Jr had an itchy trigger finger and wanted to test out the bitcoin alert system.. even he, as a pool owner and developer should know that the difficulty doesnt change for 2 weeks.. so if a new miner has come on the scene or an existing miner has upped their hashing power, the difficult wont compensate for many days..

and the chances of a miner upping their hashrate is not once in 50 years... its an every day risk..
where did luke_jr even get that 50year statistical calculation from??

[ATguy]

lets hope it will not impact the price to go down because of that

If anything this statistical anomaly is a good thing, since people will get their transactions confirmed faster while also helping to clear out the backlog in the mempool.

Now imagine the opposite statistical anomaly happens (with much less blocks found), so the backlog grows correspondingly.

What is better for Bitcoiners:

1) slowly clear the backlog over the about half day
2) flush the backlog in next block/s to blockchain

I mean, at the end of the day, the blockchain will have the same size, but with current 1) method people have unnecessary wait for their 1st confirmation many hours in average in the above scenario !

[saturn643]

this is like saying, "we had an average of 1 block every 5 minutes for the past 4 day", double of the luck, would have been an issue in case it was every 1-2 minutes maybe...

as i see it it can happen, like it happened in the past that for two hours there were no blocks

4 hours not days..

i think luke_Jr had an itchy trigger finger and wanted to test out the bitcoin alert system.. even he, as a pool owner and developer should know that the difficulty doesnt change for 2 weeks.. so if a new miner has come on the scene or an existing miner has upped their hashing power, the difficult wont compensate for many days..

This isn't the alert system, and he has no way to increase the hashpower like that unless he bought a ton of new miners to do it.

and the chances of a miner upping their hashrate is not once in 50 years... its an every day risk..
where did luke_jr even get that 50year statistical calculation from??

If you read the full quote, he is assuming that the hashrate is not increasing. That calculation is the probability of finding 1 block every 5 minutes for 4 hours.

If the network hashrate was not increasing, this event should occur only once every 50 years

[RawDog]

I am not supposed to say this, but...  (I am also not supposed to know this)

Spondoolies has just begun testing their SP50 chips.  They low-balled the performance expectations, now they are killing it.  Good news, Spondoolies is honest and not likely to be involved in any 51% bullshit.  So, everyone can relax now.  

[Erkallys]

I think this could come from BitFury expending his mining capabilities recently. Don't worry, difficulty should adjust soon to correct this. Bitcoin is indestructible !

[franky1]

looking at the last 4 difficulty change dates which should be 14 days apart (2016 blocks /144 blocks a day)

Dec 18 2015    18.14%    668,931,642 GH/s
Dec 06 2015    8.77%    566,236,898 GH/s
Nov 24 2015    10.44%    520,569,941 GH/s
Nov 11 2015    5.77%    471,360,171 GH/s

november 24th is 13 days after november 11th
december 6th is 12 days after november 24th
december 18th is 12 days after december 6th

this means that more blocks have been mined per 2weeks adjustment than expected.. meaning the difficulty calculations have not kept ahead of hashpower.

in the last 42 days there has been 5 days of extra blocks that there should not be yet.
(not saying they are invalid, just that we are ahead of the game by 5 days, atleast)
and luke_jr should know all of this and understand that its not out of the realm of reality that blocks will be solved quicker than average.

until difficulty adjustments are made to not only exceed hashrate, but also bring atleast 5 days of excess blocks into line, i dont see much that luke_jr can offer us as to a solution.

[chek2fire]

And the unconfirmed transactions drops under 2k 

[Meuh6879]

No.

http://statoshi.info/dashboard/db/transactions