4 ms·

is there potential for a denial of service attack on the blockchain if people just started inundating it with useless data entries?

bombcar 5y ago
This has basically already occurred IIRC - the entire bitcoin blockchain is now too massive for most people to entirely download and authenticate; most people rely on exchanges, etc to do it for them.
- spinny 5y ago
  Not really that big yet. i keep a full copy that is a few weeks behind, i'm running the client with the -txindex argument, which indexes all transactions (and takes up a little extra space), currently is at 414 GB. The initial sync from block 0 can take a while, but after having an almost synced copy, getting it up to date is fast I remember thinking "this will never work" after reading the bitcoin whitepaper, i assumed that at some point only supernodes would exist. We might be getting there, be we are not there yet, and definitely moving at a slower pace than my initial prediction. Holding 400-500 GB is still very attainable today at a small cost
  - sashimi-houdini 5y ago
    It only "worked" by not actually being useful as a globally distributed currency for microtransactions (which was the original promise)
retrac 5y ago
Yes, this happened very early on with Bitcoin. Casinos where every wager was a transaction. Then some started actively just spamming. It led to the transaction pricing mechanism. Every transaction includes a bid on how much you'll pay to have it included in the chain. Miners pick the highest bids from recently broadcast transactions. There's no requirement to do so but it's economically incentivized, as the miner receives the transaction fee. Any sustained attack becomes very expensive. And other users can still pay the asking price to get their transaction in the next few blocks. Bitcoin has in practice maxed out the great majority of blocks (about 1 MB per 10 minutes) for a while now, and so many cheap transactions just never get processed as they're always outbid.
jcpham2 5y ago
I did this to namecoin for fun with a bash script for 36 hours in a loop once. This is a true story. Back in the merge mining days they gave them out for free and no one knew what to do with them. I took all the public namecoin addresses I could find and put them in a text file and sent everyone dust. Fun times.
- pontifier 5y ago
  I spammed several portions of NEM and ground things almost to a halt several times. That coin sucks, and their new one probably isn't much better. I'm sure it's just a way to erase the connection to my spam and cash out a bit while changing the name. What jerks.
reidjs 5y ago
That happened on Ethereum with a game called exit.scam where the last person who buys a token wins, but every time someone buys one the timer resets. In order to win the game a bunch of miners colluded to make it impossible to get a transaction through, guaranteeing they bought the last one.

bombcar 5y ago

This has basically already occurred IIRC - the entire bitcoin blockchain is now too massive for most people to entirely download and authenticate; most people rely on exchanges, etc to do it for them.

spinny 5y ago
Not really that big yet. i keep a full copy that is a few weeks behind, i'm running the client with the -txindex argument, which indexes all transactions (and takes up a little extra space), currently is at 414 GB. The initial sync from block 0 can take a while, but after having an almost synced copy, getting it up to date is fast I remember thinking "this will never work" after reading the bitcoin whitepaper, i assumed that at some point only supernodes would exist. We might be getting there, be we are not there yet, and definitely moving at a slower pace than my initial prediction. Holding 400-500 GB is still very attainable today at a small cost
- sashimi-houdini 5y ago
  It only "worked" by not actually being useful as a globally distributed currency for microtransactions (which was the original promise)

retrac 5y ago

Yes, this happened very early on with Bitcoin. Casinos where every wager was a transaction. Then some started actively just spamming. It led to the transaction pricing mechanism. Every transaction includes a bid on how much you'll pay to have it included in the chain. Miners pick the highest bids from recently broadcast transactions. There's no requirement to do so but it's economically incentivized, as the miner receives the transaction fee. Any sustained attack becomes very expensive. And other users can still pay the asking price to get their transaction in the next few blocks. Bitcoin has in practice maxed out the great majority of blocks (about 1 MB per 10 minutes) for a while now, and so many cheap transactions just never get processed as they're always outbid.

jcpham2 5y ago

I did this to namecoin for fun with a bash script for 36 hours in a loop once. This is a true story. Back in the merge mining days they gave them out for free and no one knew what to do with them. I took all the public namecoin addresses I could find and put them in a text file and sent everyone dust. Fun times.

pontifier 5y ago
I spammed several portions of NEM and ground things almost to a halt several times. That coin sucks, and their new one probably isn't much better. I'm sure it's just a way to erase the connection to my spam and cash out a bit while changing the name. What jerks.

reidjs 5y ago

That happened on Ethereum with a game called exit.scam where the last person who buys a token wins, but every time someone buys one the timer resets. In order to win the game a bunch of miners colluded to make it impossible to get a transaction through, guaranteeing they bought the last one.