average confirmation time 18 minutes?

[C. Bergmann]

what's on?
Look at this chart https://blockchain.info/de/charts/avg-confirmation-time?timespan=1year&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

In may there was a spike in confirmation time up to 20 minutes

on november the 6th there was this spike again.

Is this just a joke of randomness?

Or do we have too much transaction for the system?

Some fork, DDoS?

[elasticband]

Difficulty out weighing hashrate for brief periods, does not happen to often as chart shows

[C. Bergmann]

Ah, difficulty jump? Or the actual drop in the hash-rate?

[DeathAndTaxes]

For whatever reasons (chance or intentional) there is a much higher than normal number of no fee and low priority txs awaiting confirmation.  Also it appears the average block size has been lower than normal.  Not sure if pools are being more conservative or it is one large miner, or just the shift in hashing power from larger block miners to smaller block miners.

Put it all together you got more transactions per second and less transactions per block.  Confirm times are going to go up.

http://blockchain.info/charts/avg-block-size?timespan=180days&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=


Still miners pools need to step up their game.  Blocks are downright tiny.  <200KB.  <100KB?  Recently there was a 47KB block (130 txs) with >5,000 tx awaiting confirmation.
It is one thing to delay/block low priority tx and spam w/ no fees.  It is another one to cap your blocks so small that hundreds of paying txs are waiting 2 or more blocks for a confirmation.

I think miners need to start having a discussion with their pool operators.  Yes larger blocks mean a small increase in orphans but most miners and pool operators should be forward looking enough to not want to kill the goose.  When Bitcoin is <1 tps and <20% of even the 1MB hard limit and thousands of tx are just idling it just is bad PR.  Everyone talks about mainstream, forget mainstream how are we going to handle even 1 tps if miners can't start clearing backlogs of txs now?

[DeathAndTaxes]

Ah, difficulty jump? Or the actual drop in the hash-rate?

Neither the time between blocks is <10 minutes.  It is just the number of transactions per 600 seconds is significantly higher than the number of tx per block.

If txs are getting added to the "unconfirmed pool" at a rate of say 0.5 tps and miners are confirming them at a rate of 0.1 tps then overtime the pool is going to get fuller and fuller.  Until either the transaction rate slows down or miners confirm more transactions the trend will continue.

For example at the time this block was mined there were >4,000 unconfirmed tx including paying tx with fees >0.1 BTC.
http://blockchain.info/block-index/438682/000000000000000412cd1343f0361f49c9f493a159ce95e72e220ca1bc3b97b5

The pool/miner decided to include only 130 tx (3% of waiting txs) and more silly only included 0.02 tx worth of fees leaving behind almost 80% of the available paying tx.  I guess this pool/miner doesn't like fees.  Maybe they are making too many BTC already?

[franky1]

miners should never only accept paid transactions EVER, but it appears that they are.

what they should do is first fill up the block with the paid transactions (prioritise) and then fill the rest of the block with free transactions waiting.

i repeat the should never only fill blocks with paid transactions.

that is selfish and does not help the overall usefulness of bitcoin. this is especially the case as the total fee's per block doesn't add up to much anyway, fee's are not really a big and noticeable bonus to the reward. so whether a transaction has a fee or not the end "reward" pot doesn't change.

take for instance
http://blockchain.info/block-index/438690/0000000000000006f53d92ffea0a2ca3f81763319b5f51498b56d60a85d25fa2
105 transactions
only 53k out of 1mb block size!!
and the end result is 0.0122101 BTC fee bonus

what makes me laugh is that the pool owner takes more of profits from all its miners then the transaction fee gives the pool. this means none of the miners ever get a bonus, above 25BTC to share between themselves.

SO MINERS, VOTE AND ENCOURAGE YOUR POOL OWNER TO FILL THE BLOCKS UP

[Lauda]

Ah, difficulty jump? Or the actual drop in the hash-rate?

Both can occur, sometimes even at the same time.

[Quantus]

How many Bitcoin miners live in the Philippines? A powerful typhoon ( the most powerful ever recorded in history) just hit Asia.

[DeathAndTaxes]

How many Bitcoin miners live in the Philippines? A powerful typhoon ( the most powerful ever recorded in history) just hit Asia.

The block frequency hasn't decreased.  The number of blocks is >6 per hour.   It isn't an issue of not enough blocks the blocks are just 90%+ empty and while normally that might be fine right now the number of transactions is high.

[bee7]

@DeathAndTaxes

might it be that the mempools are intentionally flooded with txns spending unconfirmed outputs of transactions with a very small fees?

look at this please:

http://blockchain.info/tx/e931d331b86aba238c641a90286a0490f77a96ecad9b8b3a542aa67d1544a52f
http://blockchain.info/tx-index/96182071

[Birdy]

For whatever reasons (chance or intentional) there is a much higher than normal number of no fee and low priority txs awaiting confirmation.  

Maybe one reason:
Tf sent a lot of the inputs.io refunds without fee.

[Quantus]

Maybe it's the new Chines Bitcoin users. Ether way it's time to call the major pool operators and ask them to stop being greedy bastards and let in all non fee transactions until the blocks are full.

[DeathAndTaxes]

@DeathAndTaxes

might it be that the mempools are intentionally flooded with txns spending unconfirmed outputs of transactions with a very small fees?

look at this please:

http://blockchain.info/tx/e931d331b86aba238c641a90286a0490f77a96ecad9b8b3a542aa67d1544a52f
http://blockchain.info/tx-index/96182071

Possibly but it looks like even just an avg block size of 300KB would have prevented any backlog from being created so my gut feeling is no.

[zvs]

@DeathAndTaxes

might it be that the mempools are intentionally flooded with txns spending unconfirmed outputs of transactions with a very small fees?

look at this please:

http://blockchain.info/tx/e931d331b86aba238c641a90286a0490f77a96ecad9b8b3a542aa67d1544a52f
http://blockchain.info/tx-index/96182071

the fees on those are actually standard, .0001 per 1kb ... which I think is too low.  maybe that's the same feeling from some pool ops.  i had mine set to only accept 0.0005 or higher for a long time (months past the "standard" client change from .0005 to .0001)

[Leehoya]

Too much selfish miners out there. Nearly 24 hours and havent been in a block. Thanks to Blockchain android wallet. https://blockchain.info/tx/b9163c219f68cb14aed197d436f30e7e8b1695bba75d4310e1455fa2f902e992

[laowai80]

Maybe it's the new Chines Bitcoin users. Ether way it's time to call the major pool operators and ask them to stop being greedy bastards and let in all non fee transactions until the blocks are full.

So, how are miners greedy and users who don't want to send them a tiny 0.0001 (3 cents) tx fee not greedy?

All participants need to responsibly treat the network and be nice to each other. Without miners bitcoin is dead, why not pay them this small fee to keep the network running? Truely is this the "why worry about a speck in your friend's eye when you have a log in your own" kind of thing.

Somebody always has to pay for the banquet, there is no free cheese ever. So grow up, and understand you have to pay for everything in life, even in bitcoin. Bitcoin is run and used by people, and people are always sinful, it's their nature, so pay your dues and enjoy the system.

[Lauda]

Sometimes it's faster and sometimes it's slower. That's life.

[MaxBTC1]

The other thing is, surely as less mine it will take longer to confirm transactions

[Lauda]

The other thing is, surely as less mine it will take longer to confirm transactions

Actually more and more people are mining.

[DeathAndTaxes]

The other thing is, surely as less mine it will take longer to confirm transactions

No.  The network is self adjusting. 1 person (or more importantly 1 GH/s) or 1 million people (or 1 quadrillion GH/s) the network will target 10 minutes per block.