2015: The year of the stress test

[worhiper_-_]

You should ask Amaclin https://bitcointalk.org/index.php?action=profile;u=197593 about this.

https://bitcointalk.org/index.php?topic=1198032.0
https://bitcointalk.org/index.php?topic=1166928.0

Is amaclin's "Cheap way to attack blockchain" what's currently used to stress test bitcoin?

no

Oh, thanks for the clarification. That seems a bit surprising though, seeing how cheap and efficient the method you detailed would be for a 'stress test'.

Edit: now that I think about it, the current stress test likely has the goal to crash nodes. It's not the most efficient block filling attack but from what we see it was quite effective at crashing nodes. Would the SIGOP stress test have any success at such a thing?

[amaclin]

Oh, thanks for the clarification. That seems a bit surprising though, seeing how cheap and efficient the method you detailed would be for a 'stress test'.

I do not want to spend even 0.1 btc for stress-testing even with "cheap" method
I am not so rich as coinwallet.eu
 

[Meuh6879]

I would love to hear the opinions of a p2p network specialist on the block size issue.  

  first, not any P2P software store mempool (cache) on RAM more than 1 minute : Bitcoin Core store this indefinitly = crash of the node when we have SPAM.

second, P2P software use drive in better way than Bitcoin Core does ... what a hell is this software that it must read 4,2Go of files with a 200Mb of cache ?!?

three, why this "node" must to use more than 20% on a 3GHz system (2 cores) ? just to store and compare transactions ? Come one ... hashset organisation do this more quickly AND more sanity (less freezing in the environment of others running programs) than DBberkeley organisation.

i understand why, now, the developper don't love the architecture of the build bitcoin core genesis code ...

---

But i understand the thing when developpers says : i don't run a local node ... i loan a server in a datacenter instead.

  it's not a real and local P2P solution, now ... it must have ressource and power, now.

Well, i regrete this but it's internet money, too.
I prefer rigurous money supply (finished supply) than the actual mess of the economies (you work, you have money and then, a chimpanzee in a central bank emit money = time work value destroyed).

[amaclin]

Edit: now that I think about it, the current stress test likely has the goal to crash nodes. It's not the most efficient block filling attack but from what we see it was quite effective at crashing nodes. Would the SIGOP stress test have any success at such a thing?

SIGOPS-attack can be used for doublespending and delaying regular users' transactions up to several days even the fees are sufficient

[tommorisonwebdesign]

This stress test may be a good thing for Bitcoin. It can show that the network can handle a large load and all in all it may raise the price. I think that we actually may have something to gain from all of this.

[Raimonn]

Its nice to see that bitcoin is good enough to pass all the stress tests that someones are doing this year.

[alani123]

ELI5: How would the SIGOP attack be used for double spending?

[aso118]

2015 is the year of the stress tests and the block size debate.
I think the block size debate is the cause and that is what 2015 will be remembered by.

[shorena]

-snip-
Umm... Anyway the bottom line is the network will be fine. Weaker nodes like the raspberry pi's and old computers' may crash or run out of Ram forcing them to use disk space thus slowing them down and potentially increasing propagation time accross the network, leading to a more fractured network and more orphaned blocks causing even more bloat on full nodes that have to store all these orphaned blocks but having said all that everything will be fine. Really!  

What's the maximum amount of RAM a top of the range node is likely to be capable of having? I have seen bog standard computers with 8GB, but servers are designed to allow a much higher maximum, if needed. I'm not familiar with the maximum top of the range ASICs can use, or if they are designed like servers so their RAM can be upgraded to a very high amount.

Durring the recent spam my node crashed after ~4 hours because it ran out of memory. Its using 2 GB and I currently have to filter TX by fee in the hopes that another such attack will not result in a crash.

I also considering a server upgrade to 4GB which would have dealt with this attack w/o issues, but might still result in issues with a larger attack. Under normal circumstances (as they are now) it hovers around 900MB RAM used system wide.

[Meuh6879]

I currently have to filter TX by fee

solution : https://medium.com/@octskyward/mempool-size-limiting-a3f604b72a4a

work great, no fees filter applied during this SPAM test.

[gogxmagog]

I don't really understand the motivations for some of this stuff, but it is important that it happens. It won't get anywhere unless it's given a beating and comes out the other side the stronger for it.

im guessing its real hard core hacker culture. theyre doing it for bragging rights and to prove they can.

and yes, it helps in the long run.

of course the mainstream media will have a hey day with all the FUD they can proliferate about it... well be hearing about it for years. just like gox, silkroad, all the dark web scare propaganda. it does hinder adoption somewhat (hopefully temporarily)

[notbatman]

Well what ever they're doing today it seems to be having a minor effect on my node. My hash rate is down almost 1% and my miner temperature is up 2°F.

That's some serious bragging rights there. 

[Meuh6879]

well, it's true ... but, setting provide by XT (max transaction in mempool) work well now.

[notbatman]

well, it's true ... but, setting provide by XT (max transaction in mempool) work well now.

Heh, are you running a BFL Jalapeno or is that a Little Single?

[Meuh6879]

R-Box (same building system than U3).
It's to view the activity inside of Bitcoin Network (server mode) ... not to really mine.

[Q7]

First coinwallet.eu, then malleability and now a powerful stress test from an unknown source comes in creating a growing backlog of ~360Mb

I didn't know another "test" was on until I read your post. I check the explorers, we no have 28k unconfirmed transaction totaling 310M. The good news is nothing a 0.0002btc/k fee wouldn't solve.

The part I don't understand is why are some people burning their bitcoin to repeat these tests? They haven't proven anything and it's not bringing the network to its knees. Why waste money?

Which I have a feeling the money actually comes from a source where the people there want to see bitcoin failed. And one more thing we shouldn't forget is that fiat can be printed infinitely so whatever funds can actually be generated on the go. Even if the objective does not succeed to prove anything, the money is still basically free money.

[aso118]

I don't really understand the motivations for some of this stuff, but it is important that it happens. It won't get anywhere unless it's given a beating and comes out the other side the stronger for it.

im guessing its real hard core hacker culture. theyre doing it for bragging rights and to prove they can.
and yes, it helps in the long run.
of course the mainstream media will have a hey day with all the FUD they can proliferate about it... well be hearing about it for years. just like gox, silkroad, all the dark web scare propaganda. it does hinder adoption somewhat (hopefully temporarily)

It really doesn't seem to have an impact on the price of bitcoin; although you would expect some downward pressure.
So there is no motivation for large traders to play a role in this.

[designerusa]

Bitcoin passed this stress test this month. Its price went up again after bad days.
Some people wanna manipulate heavy use of bitcoin but they always fail.

[BitmoreCoin]

Bitcoin passed this stress test this month. Its price went up again after bad days.
Some people wanna manipulate heavy use of bitcoin but they always fail.

Those stress tests proved the resilience of Bitcoin, it is actually good for bitcoin. So the great days are ahead, indicated by the recent rise of bitcoin price.

[BlackPanda]

Bitcoin passed this stress test this month. Its price went up again after bad days.
Some people wanna manipulate heavy use of bitcoin but they always fail.

I believe bitcoin will rise again next month. This month we've been able to get the standard price for bitcoin. it all will continue to rise.