Topic: Next level Bitcoin stress test -- June 29-30 13:00 GMT 2015 - page 5. (Read 16077 times)

So has the mempool ever before been as large as it was yesterday (21Mb) ?

I really enjoyed observing how the network reacted to this increased 'stress'

jbreher, I hope you will perform a test as well. These are very important, to find weaknesses in BTC companies' implementations, etc

It's amazing to me how many of the posts in this thread are just shouting "Luke shouldn't be doing this!!!".  As if what they shout somehow matters.  People need to listen to the miners who are dropping facts on you.  They keep tellling you "I am a miner and I decide what to do with my hashrate.  I decide what to try to process and what to ignore".  The point of this setup is that everyone is free to implement their own mining policy and see how it goes for them.  For people claiming that no-fee transactions should be blocked: consider that as long there's a block reward I can set up my miner to only hash no-fee transactions, as long as I find a block I get the block reward and that far-outweighs the fees associated with a given block (on average, nowadays).

I, personally, think it was very noble of Eligius pool to ignore "stress-test" transactions, this was an example of a mining pool putting the interests of the regular, everyday, bitcoin users above their own profit interests.  This is very interesting because miners and users are often seen as having opposing interests.

Anyway, consider please, folks, shouting about what miners should do won't change the fact that they are free to do as they please.  This is the nature of a trustless, distributed, decentralized system.  The system provides conditions under which everyone, while working for their own interest, can secure the network for all.  If you work against your own interest (ie, say, you spend 20BTC just to spam for fun) you are free to do that.  Along the same lines miners are free to ignore higher-fee transactions of said spammer. 

Them's the facts!

Well, I was loose with my terminology..  Dust and No fee I don't consider valid.

The point is that normal fee paying transactions were blocked.

Wow!  That was quite a change from

You are of course entitled to your own opinion on mining policy, just as all the miners, users and even non-users.  I don't think any pools or solo miners block normal valid transactions that pays a fee.  Most allow no fee transactions as well, as long as there is room in the block and the priority is high enough.

If I had the hashrate to make a difference, I would give a high priority to transactions which reduce the UTXO set significantly, e.g. by combining a lot of dust outputs to one or two more sensible sized coins.  I would mine those transactions for free if the gain was high enough, since it saves resources, and therefore reduces cost, in the long run.

Unconfirmed transactions left, right and center. If this isn't proof that we need a blocksize increment then I dont know what the hell it is. LN is not a solution long term.

How on earth are you going to accomplish that, even if you were all the governments on Earth?  You could give all the orders you want.  It would be impossible to separate transactions from Greece from the rest.  Bitcoin doesn't know geographical boundaries.

Dust txn. Fine (to block).

No fee txn. Fine (I don't think zero fee txns should be allowed at all actually).

'Spam' in all it's varied forms. Fine.. (Well, until some size fix solution is found. Then if you pay, and there's room, you should be allowed through.)

But a normal valid txn that pays the standard fee. That should be allowed through.

Miners should LOVE to mine txns that pay fees. As Luke himself said, these are numbers, not people. There should be NO POLITICS involved. Otherwise, it's a very slippery slope.

You should stop using Bitcoin Core then, or any P2P based coin with users or miners.

Bitcoin Core refuses to make, forward or mine most valid transactions.  The set of valid transactions it accepts by default is only a fraction of all valid transactions.  It won't make, forward or mine transactions with dust outputs, for example.  This is a very effective measure against spam, and against DoS by exploding the UTXO set.  All users can change this policy if they want to, and even set it to accept all valid transactions if they have enough RAM and CPU to accept all the junk transactions which is never going to get mined.  You could even keep them in your mempool and rebroadcast them forever.  My definition of dust is 10 times larger than the default definition.  This saves space in my mempool, and fortunately you can't force me to waste my resources on your junk transactions.  This is my choice when I choose to spend my resources on running a full node.  You are free to make your own choices.

Miners can decide which transactions to put in their blocks.  This is true for all coins which use distributed mining.  Some pools use stricter than default settings.  Most use the default settings.  Some, particularly Eligius, accept most valid transactions which other pools will refuse to mine.  Eligius demands a fee for that kind of transactions, which is their choice.  Some, e.g. Eligius, will let you add a fee to a transaction by sending the change output with a higher fee, considering the fees and sizes of both the parent and the child when prioritizing transactions.  Most pools will only consider the fee of the parent transaction.  This is entirely up to them.

Finally got my BTC. What a nightmare, I'm glad it was only a relatively small amount, I would have been really anxious if I was waiting on 10+ BTC. Last night was the first time I've ever had a transaction going through when there was a stress test. I'm not a tech expert so what did they actually learn from spamming the network & making lots of people worry about non confirmed transactions.
I checked bc.info & at one point there were over 11,000 unconfirmed transactions. Was it really necessary for that many?

Greek people will use VPN instead 

Actually I don't know if you can specifically broadcast your transactions to one pool that you prefer. If pools have different policies when it comes to transaction priority, then maybe 0 fee transactions could still be done with some pools (They block spam transactions intentionally). And overtime, more and more spam transactions will be queued in the memory pool due to more and more pools are blocking them

Then we have an interesting situation that although there are 200MB queued transactions in memory pool, most of them are spam transactions

Blocking any transactions is the thin end of the wedge. If it starts for any reason it will snowball until international governments start ordering pools to block a regularly updated list of transactions. What if the pools were ordered to stop processing transactions from Greece?

I have to completely agree with Kano here..

A valid TXN is a VALID txn. Full stop. End of story. NO POLITICS.

If someone starts to show that it is possible to 'edit' which transactions go through the system, then I think it sets a very bad precedent.

I don't see why the miners weren't really happy to be slurping up all those extra fees !? That's what they live for.. isn't it.. ?

FREE BUFFET!! EAT AS MUCH AS YOU LIKE!! 

(Ok - so we made you work for it a little harder today, but we don't want you getting fat and lazy..)

Do you oppose the default anti-DoS filters which are built into Bitcoin Core as well, or just this particular anti-DoS filter because of problems with a specific person?

Perhaps you and I have different ideas of what the problem is?  To me the problem is the fact that legitimate low fee transactions were replaced by malicious spam.  Eligius solved this by allowing even the low fee transactions confirm, in the same way as the default settings in Bitcoin Core filter spammy transactions which used to cause DoS earlier.

Whatev. I'll run my own damn test. At the same time, I'll generate a transaction here and there, with some nominal increment over the standard fee you're paying. I bet I spend less, and end up more satisfied in the outcome, than do you.

100% ... no that's false and you know it is.

Doing Luke's bidding? Not gonna happen by anyone but you.

The simple fact is that they were valid transactions with fees.

If following Luke's orders, your pool decides it wants to block valid transactions, that's up to him.

However, as I stated:
In fact what Eligius did, helped prolong the problem.

Not sure where you get your facts from, but as I had mentioned previously in this thread, there was no "blacklist" in play with regard to filtering this particular spam attack.  The transactions are pretty easily detectable because the attackers really don't seem to know what they're doing.

The only things prolonging the attack are the attackers and other pools/miners *not* filtering the scammer's spam.  The other pools/miners would rather fill their blocks with the attacker's transactions for an extra 0.1 BTC in fees vs protect bitcoin as a whole... or it's laziness.  Either way, any ill effects of this attack could easily have been countered 100% with the participation of just a few of the larger pools.

I prefer fungibility, which is the basic concept of Bitcoin, the peer to peer digital currency, not blacklisting.

However, Eligius' blacklisting had no positive effect on the stress test and helped no one.
In fact what Eligius did, helped prolong the problem.

Thanks for the support!  It is appreciated.

Me too... I supposed to have my payout from my sig. campaign few hours ago, however it is still unconfirmed until I've sent my transaction with extra fees (I've paid 20000 satoshis for just a 3xx byte tx!) and the sig. campaign's payment and my payment confirms in the same block, with my tx just after the sig. campaign's tx (so all the participants should thank me for paying extra fees to let their tx to be confirmed ). I think such tests should stop immediately, and none of them should be carried anymore. Please stop spamming the blockchain.

I had to move some miners out of BTC Guild at the last minute before it shut down and picked Eligius because of this. Call it filtering, blacklisting or whatever ... I think that this response was appropriate.