Stress test incoming and an update on blocksize

Thursday 10 September 2015

TL;DR Expect delays in the coming days - raise your fees to make sure txs confirm fast

A new stress test is expected to hit the bitcoin network and cause significant delays starting today, September 10, 10am GMT.

Run by CoinWallet, the ‘test’ is supposed to measure how the bitcoin network reacts to a large number of very small transactions. However, critics effectively view it as a DDoS (distributed denial of service) attack, which is what it is: the huge number of transactions clog the queue to be included in the next block, meaning that ‘legitimate’ txs may be delayed by many hours (last time around it took me 12 hours to see a single confirmation on an otherwise normal transaction). You can improve your chances of getting a tx into the next block by raising the fees paid.

Read also: Clogged - the anatomy of a bitcoin attack


CoinWallet: happily bringing us something no one wants

CoinWallet’s so-called test isn’t just inconvenient - and, frankly, selfish, particularly given that this isn’t the first time. It’s an act of propaganda that plays into the whole blocksize debate - since larger blocks are more expensive to fill up with junk and therefore the network is less likely to be DDoS’d.

On which.

XT proves unpopular

Gavin Andresen’s XT solution, which implements his Bitcoin Improvement Proposal (BIP) 101 - the increase to 8MB blocks, doubling every two years thereafter - has failed to win significant support. This is despite the fact that BIP 101 was agreed after consultation with a cartel of Chinese mining pools, which together account for some 60 percent of hashrate.

The problem? XT controversial. Raising the blocksize was controversial anyway, but the inclusion of some unexpected code that could be used to block particular IP addresses and therefore create blacklists - killing bitcoin’s fungibility and introducing a nasty element of censorship - has incensed critics. The official explanation is that the code is required to prevent DDoS attacks on the network, which might be fair enough, but this was clearly not the time to throw any more fuel on the fire. Moreover, XT was created without consensus among bitcoin’s core developers. It's a unilateral fix. Consequently, it’s viewed with some degree of suspicion.

So it is that only a handful of nodes - around 12 percent - have opted to go for Gavin’s XT solution. BIP 101 might still be incorporated into the bitcoin Core. The alternative, which is winning more support from miners, is Jeff Garzik’s BIP 100. You can see who's mining what here.

BIP 100 is somewhat more complicated than BIP 101, but essentially it will allow miners to vote for the blocksize they want, with limits to the rate of change allowed and to the ultimate blocksize, which is capped at 32 MB (rather than XT’s final cap of 32 GB). You can read a more detailed explanation here

It’s not perfect, by any stretch, but it's certainly more democratic. Miners are backing it to the tune of around 60 percent.

The saga will no doubt continue - but the sooner it ends, the better for everyone involved in bitcoin.


comments powered by Disqus