CoinWallet says Bitcoin stress test in September will create 30-day backlog

[CohibAA]

So we need to see whatever meager fee the attacker is paying, and set this minrelaytxfee slightly above and thus we make the attack ineffective.

If I remember correctly, at some point during the last "test" they began to increase the fees on the dust transactions to ensure they would continue to propagate.

[ChetnotAtkins]

Yes I think I remember they even used standard fees, at least for some tx

[neochiny]

So we need to see whatever meager fee the attacker is paying, and set this minrelaytxfee slightly above and thus we make the attack ineffective.

If I remember correctly, at some point during the last "test" they began to increase the fees on the dust transactions to ensure they would continue to propagate.

so that means that we dont have to worry anything if we just pay that... thats the conclusion of all of this?

[Cryddit]

This "stress test" is, IMO, probably being done by a coalition of mining pools.  It exists for the plain and simple purpose of increasing mining profits by forcing users to pay higher fees.

Assuming they are X% of the hashing power, they are getting X% of the fees they put into it back when they get blocks.  So the expense is less than it appears to be.  They make a profit to cover the (100 - X)% because they ALSO get X% of the increased fees that users are putting in to cut ahead of their spam flood.

If we assume that legitimate users will pay increased fees for half the block space, and that the coalition represents more than 38% of the mining power, then spamming the network to keep blocks half full of spam (leaving the other half available for increased fee prices) is more profitable to them than not doing so.  

In practice it's even more profitable than that (or becomes profitable even for a coalition smaller than 38%) because (a) users outbidding them for block space are setting their fees to a nontrivial amount *above* the fee level of the spam rather than just equal to it, and (b) lots of users leave their fee settings higher even when the spammers spend a few weeks not supporting the increased cost of the "stress test".

I'm betting the people doing it DON'T want a switch to Bitcoin-XT; with the larger block size of Bitcoin-XT, legitimate traffic willing to pay the premium to cut ahead of the spam would be less than 1/2 the maximum block size and it wouldn't be profitable any more.

[ChetnotAtkins]

You are wrong. Unsurprisingly the Coinwallet spam is conducted by XT shills:

“We feel that our tests might prove to be the catalyst that propels the core devs and miners to implement the required hard fork that is desperately needed,” CoinWallet mentioned.

Link: http://www.newsbtc.com/2015/08/19/coinwallet-september-bitcoin-stress-test-could-create-backlog/


Just one of the ugly strategies involved in this coordinated attack against Bitcoin. I don't care about Blockstream either but we have to stop XT

[Cryddit]

You are wrong. Unsurprisingly the Coinwallet spam is conducted by XT shills:

“We feel that our tests might prove to be the catalyst that propels the core devs and miners to implement the required hard fork that is desperately needed,” CoinWallet mentioned.

And you expect that they're being honest about why they're doing it.  I think that makes you what the cons used to call an "all-day sucker."

[Carlton Banks]

This "stress test" is, IMO, probably being done by a coalition of mining pools.  It exists for the plain and simple purpose of increasing mining profits by forcing users to pay higher fees.

I'm betting the people doing it DON'T want a switch to Bitcoin-XT; with the larger block size of Bitcoin-XT, legitimate traffic willing to pay the premium to cut ahead of the spam would be less than 1/2 the maximum block size and it wouldn't be profitable any more.

Why would it be any more plausible that pro-Core miners are planning this than that pro-XT miners are? Both stand to gain equally from doing so, and both also stand to gain from firstly conducting the attack, then secondly blaming the other party to discredit them.

[MoorChael]

Bitcoin is quite popular and it may not change a lot for that.

[Cryddit]

Why would it be any more plausible that pro-Core miners are planning this than that pro-XT miners are? Both stand to gain equally from doing so, and both also stand to gain from firstly conducting the attack, then secondly blaming the other party to discredit them.

Ding!  It's all about the profit motive.  Regardless of whether the miners claim tobe Pro-XT or Pro-Core, some coalition of them is making money by doing these "stress tests" that won't work under XT. 

[Carlton Banks]

Why would it be any more plausible that pro-Core miners are planning this than that pro-XT miners are? Both stand to gain equally from doing so, and both also stand to gain from firstly conducting the attack, then secondly blaming the other party to discredit them.

Ding!  It's all about the profit motive.  Regardless of whether the miners claim tobe Pro-XT or Pro-Core, some coalition of them is making money by doing these "stress tests" that won't work under XT.  

No, that's incorrect.

XT miners have the same opportunities to collect fees as 1 MB miners, given a transaction rate that fills an equivalent proportion of a given block size. Is that not what this debate is about, increasing the carrying capacity?

You cannot make both arguments simultaneously; that tx rates will remain at present day levels post-fork and the miners will lose profitability, and yet also that transaction rates are set to soar which you suggest is not motivating miners at all? Are you making an honest argument Ray Dillinger, or did you just overlook that particular glaring contradiction?

[hdbuck]

This is not a test, it's called sabotage. They want to push people to pay higher transaction fee's and that has no benefit for the Bitcoin users.

The last "test" cost +/- $5000 per day, so they should in theory make more than that to break even.

The problem with these "attacks" is that other people would piggyback onto this and then achieve a much better result with their hostile attack, than what they would have achieved, when they did it alone. It would

become much cheaper to bring down the network.

it is to push to block increase agenda.

[Mickeyb]

who cares.  If bitcoin can't stand a stress test then it will never make it as a global payments protocol.

Well that's why we need bigger blocks. I bet that stress tests that were going on lately have been and the future ones as well are here only to show us that the bigger blocks are the necessity.

The thing is also that we can't prevent these stress tests. People can be doing them as much as they want, all they need is some cash. And whole Bitcoin ecosystem is suffering during these stress tests.

[Carlton Banks]

The thing is also that we can't prevent these stress tests. People can be doing them as much as they want, all they need is some cash. And whole Bitcoin ecosystem is suffering during these stress tests.

Wrong. Core team have been working on a solution/mitigator to these attacks for several weeks already. It will make it into version 0.11.1 or 0.12, depending on how Wladimir decides to do it.  

[Mickeyb]

The thing is also that we can't prevent these stress tests. People can be doing them as much as they want, all they need is some cash. And whole Bitcoin ecosystem is suffering during these stress tests.

Wrong. Core team have been working on a solution/mitigator to these attacks for several weeks already. It will make it into version 0.11.1 or 0.12, depending on how Wladimir decides to do it.  

Oh ok, I didn't know that! Can I get a link to learn more about it.
Thanks!

[Carlton Banks]

The thing is also that we can't prevent these stress tests. People can be doing them as much as they want, all they need is some cash. And whole Bitcoin ecosystem is suffering during these stress tests.

Wrong. Core team have been working on a solution/mitigator to these attacks for several weeks already. It will make it into version 0.11.1 or 0.12, depending on how Wladimir decides to do it.  

Oh ok, I didn't know that! Can I get a link to learn more about it.
Thanks!

It's all code and discussions about the code at this point, but here's what I saw: https://github.com/bitcoin/bitcoin/pull/6557

Check out the pull request referenced in the first post, I think that's the original design that gained acceptance.

[valiz]

+    strUsage += HelpMessageOpt("-maxmempool=<n>", strprintf(_("Keep the transaction memory pool below <n> megabytes (default: %u)"), DEFAULT_MAX_MEMPOOL_SIZE));
+    strUsage += HelpMessageOpt("-mempoolexpiry=<n>", strprintf(_("Do not keep transactions in the mempool longer than <n> hours (default: %u)"), DEFAULT_MEMPOOL_EXPIRY));
+    strUsage += HelpMessageOpt("-limitancestorcount=<n>", strprintf(_("Do not accept transactions if number of in-mempool ancestors is <n> or more (default: %u)"), DEFAULT_ANCESTOR_LIMIT));
+    strUsage += HelpMessageOpt("-limitancestorsize=<n>", strprintf(_("Do not accept transactions whose size with all in-mempool ancestors exceeds <n> kilobytes (default: %u)"), DEFAULT_ANCESTOR_SIZE_LIMIT));
+    strUsage += HelpMessageOpt("-limitdescendantcount=<n>", strprintf(_("Do not accept transactions if any ancestor would have <n> or more in-mempool descendants (default: %u)"), DEFAULT_DESCENDANT_LIMIT));
+    strUsage += HelpMessageOpt("-limitdescendantsize=<n>", _("Do not accept transactions if any ancestor would have more than <n> kilobytes of in-mempool descendants (default: -maxmempool*1000/200). Increases to this value should be made by increasing the value passed to -maxmempool"));

+/** Default for -maxmempool, maximum megabytes of mempool memory usage */
+static const unsigned int DEFAULT_MAX_MEMPOOL_SIZE = 500;
+/** Default for -mempoolexpiry, expiration time for mempool transactions in hours */
+static const unsigned int DEFAULT_MEMPOOL_EXPIRY = 168;
+/** Default for -limitancestorcount, max number of in-mempool ancestors */
+static const unsigned int DEFAULT_ANCESTOR_LIMIT = 100;
+/** Default for -limitancestorsize, maximum kilobytes of tx + all in-mempool ancestors */
+static const unsigned int DEFAULT_ANCESTOR_SIZE_LIMIT = 1000;
+/** Default for -limitdescendantcount, max number of in-mempool descendants */
+static const unsigned int DEFAULT_DESCENDANT_LIMIT = 1000;

Down with stress tests spam attacks! 

[ChetnotAtkins]

You are wrong. Unsurprisingly the Coinwallet spam is conducted by XT shills:

“We feel that our tests might prove to be the catalyst that propels the core devs and miners to implement the required hard fork that is desperately needed,” CoinWallet mentioned.

And you expect that they're being honest about why they're doing it.  I think that makes you what the cons used to call an "all-day sucker."

Do you also believe Gavin and Mike are lying or are you dishonestly opportunistic with your assessment?

Surprisingly another XT shill that resorts to evasiveness when confronted with factual statements.



So why don't you tell us all here, why you are pushing vehemently for a damaging hardfork to occur? What are your motives?

Edit: found his motive for XT shilling, he is an altcoin developer. How surprising!

[galbros]

Assuming they are X% of the hashing power, they are getting X% of the fees they put into it back when they get blocks.  So the expense is less than it appears to be.  They make a profit to cover the (100 - X)% because they ALSO get X% of the increased fees that users are putting in to cut ahead of their spam flood.

I think this is a great point.  The cost to them is partially offset. 

Kind of sad that some segments of the bitcoin community would act so opportunistically against the group as a whole.  Normally, I am a fan of freedom and self interest but this seems counter productive in the long term as it reduces the attractiveness of using bitcoin and risks segmenting the community as we see with the XT-Core feud.

[Carlton Banks]

Kind of sad that some segments of the bitcoin community would act so opportunistically against the group as a whole. 

That implies you know that miners are responsible for the attack, which is not true. There exists no evidence that indicates the perpetrator of the spam attacks; none. Everything is speculation at this point.

[Cryddit]

Ding!  It's all about the profit motive.  Regardless of whether the miners claim tobe Pro-XT or Pro-Core, some coalition of them is making money by doing these "stress tests" that won't work under XT.  

No, that's incorrect.

XT miners have the same opportunities to collect fees as 1 MB miners, given a transaction rate that fills an equivalent proportion of a given block size. Is that not what this debate is about, increasing the carrying capacity?

You cannot make both arguments simultaneously; that tx rates will remain at present day levels post-fork and the miners will lose profitability, and yet also that transaction rates are set to soar which you suggest is not motivating miners at all? Are you making an honest argument Ray Dillinger, or did you just overlook that particular glaring contradiction?

The "stress test" ceases to be profitable to a mining coalition if it costs them more to drive fees up than they get in increased fees.

Increasing the block size doesn't affect the opportunity to collect fees; it does however affect the amount of tx volume a coalition would have to pay fees for before it would have the effect of driving up costs for users and fee revenue for miners.

It increases the expense of the attack.