Let's welcome the stress test

[RGBKey]

Nice try online wallet PR team /s

You write some good points though, it will be interesting to see how it pans out.

[boopy265420]

This was another unexpected stress test which made my mind up to say good bye to blockchain for good.I  woke up today and wanted to transfer some funds and it took all the day long to be up again.All this uncertainty will have no good impact on blockchain users.

[buddu]

Stress test suck, they delay a lot of my working schedule, and it is not fair for small users. There might be a need to do one every year or so, but to have Bitcoin malfunctioning for days really does harm.

No one could actually do something about it, everyone can do what they want to do in the network as there would be no regulations to stop them. With every stress-test, just think of something like this: someone is burning loads of money in order to push the network to its limit and strengthen it by doing so.

Lets be hopeful and positive to not see these things happening in future.I think in future there will be more legislation and protection for Bitcoin users and we will not see at least unseen down times on name of stress test.They need to compensate the affected users.

[Mickeyb]

Stress test suck, they delay a lot of my working schedule, and it is not fair for small users. There might be a need to do one every year or so, but to have Bitcoin malfunctioning for days really does harm.

No one could actually do something about it, everyone can do what they want to do in the network as there would be no regulations to stop them. With every stress-test, just think of something like this: someone is burning loads of money in order to push the network to its limit and strengthen it by doing so.

Weren't some devs suggested certain improvements to prevent these kind of stress tests. During the last stress test, I I have written the same post as yours, that we can't do anything about. But then one of the hero members has written that measures have taken places to prevent such stress tests in the future. I don't know is something like this possible, maybe somebody here has more info.

[johnyj]

Stress test suck, they delay a lot of my working schedule, and it is not fair for small users. There might be a need to do one every year or so, but to have Bitcoin malfunctioning for days really does harm.

No one could actually do something about it, everyone can do what they want to do in the network as there would be no regulations to stop them. With every stress-test, just think of something like this: someone is burning loads of money in order to push the network to its limit and strengthen it by doing so.

Weren't some devs suggested certain improvements to prevent these kind of stress tests. During the last stress test, I I have written the same post as yours, that we can't do anything about. But then one of the hero members has written that measures have taken places to prevent such stress tests in the future. I don't know is something like this possible, maybe somebody here has more info.

There are simple ways to filter out the spam transactions, but I'd rather not do it, just let the traffic fill the mempool and stress the nodes so that we can have a rough estimation what kind of load there will be if we use 8MB blocks (A total crash of almost every node maybe?)

The stress test is a good simulation of future when every block is full, we will see what kind of solution people will come up with when they experience delay in transaction

A typically reaction is to wait and combine several small transactions into a large one and send it with a higher fee (just like ATM cash withdraw in foreign country, you always do the maximum withdraw to reduce the impact of the fee since fee is fixed regardless of withdraw amount)

Then maybe rely on some online web wallet service since they typically combine lots of transactions into one thus can afford much larger fee

Anyway, currently the fee is really small, I guess most of the users will just raise the fee by 10x and be happy with it

Humans are forgiving

[countryfree]

A typically reaction is to wait and combine several small transactions into a large one and send it with a higher fee (just like ATM cash withdraw in foreign country, you always do the maximum withdraw to reduce the impact of the fee since fee is fixed regardless of withdraw amount)

That depends on the bank. Usually, there are 2 fees. One from the bank which issued the credit card, and another from the foreign which provide the cash. One can be fixed, the other can be %, there are many cases...

Regarding the stress tests, I welcome them. They show what it will be like next year, with more people using BTC. Banks also do stress tests!

[MicroGuy]

A typically reaction is to wait and combine several small transactions into a large one and send it with a higher fee (just like ATM cash withdraw in foreign country, you always do the maximum withdraw to reduce the impact of the fee since fee is fixed regardless of withdraw amount)

That depends on the bank. Usually, there are 2 fees. One from the bank which issued the credit card, and another from the foreign which provide the cash. One can be fixed, the other can be %, there are many cases...

Regarding the stress tests, I welcome them. They show what it will be like next year, with more people using BTC. Banks also do stress tests!

Stress tests don't have to be disruptive. That's why we have testnet.

This isn't a 'stress test', it's a premeditated attack on the network.

[tl121]

These stress tests only exist and have relevance because of the lack of leadership in the bitcoin community.   What is being tested is not the bitcoin protocol or software.  What is being tested is the  possibility of a useful community without competent leadership.

It is obvious that none of these stress tests would be happening if the blocksize issue had been resolved.  The blocksize issue has been festering for years. Solving this problem is basically a one line code change.  The inability of the so-called "core" developers to resolve this problem demonstrates their total lack of leadership skills.

[johnyj]

A typically reaction is to wait and combine several small transactions into a large one and send it with a higher fee (just like ATM cash withdraw in foreign country, you always do the maximum withdraw to reduce the impact of the fee since fee is fixed regardless of withdraw amount)

That depends on the bank. Usually, there are 2 fees. One from the bank which issued the credit card, and another from the foreign which provide the cash. One can be fixed, the other can be %, there are many cases...

Regarding the stress tests, I welcome them. They show what it will be like next year, with more people using BTC. Banks also do stress tests!

Stress tests don't have to be disruptive. That's why we have testnet.

This isn't a 'stress test', it's a premeditated attack on the network.

A stress test on testnet is not going to give you the most important feedback from all the actors in the whole ecosystem. You can not anticipate what each users will do, each of them will come up with their own solution to deal with the transaction difficulty

[johnyj]

These stress tests only exist and have relevance because of the lack of leadership in the bitcoin community.   What is being tested is not the bitcoin protocol or software.  What is being tested is the  possibility of a useful community without competent leadership.

It is obvious that none of these stress tests would be happening if the blocksize issue had been resolved.  The blocksize issue has been festering for years. Solving this problem is basically a one line code change.  The inability of the so-called "core" developers to resolve this problem demonstrates their total lack of leadership skills.

There is no leadership in a decentralized society, only the best effort to reach consensus. Luckily now we have internet and advanced communication tools like twitter/reddit/forum to quickly exchange ideas, it is possible to reach consensus without leadership

The difficulty in deciding block size just showed that consensus has not reached, without consensus, there will be no change. If every block is full and many people are complaining then there will be a consensus to raise the block size

[tmltd]

A stress test on testnet is not going to give you the most important feedback from all the actors in the whole ecosystem. You can not anticipate what each users will do, each of them will come up with their own solution to deal with the transaction difficulty

I hope automobile manufactures don't replace their dummies with real people to get better feedback...  

In all fields of science the tests are done on simulated environments. If coinwallet.eu really wanted to test and not to attack the bitcoin network and the testnet is not good for them then it ought build a simulated environment with nodes, miners, users doing tx in a big scale using cloud technologies but that needs serious money and expertise.

[tvbcof]

Satoshi-dice was sort of the stress-test of it's day.  It was annoying, but ultimately I think it was one of the best things that happened for the system since it spurred a range of developments (code and otherwise.)  I said so at the time IIRC.  Overcoming that challenge gave me and I suspect other people more confidence in the system.

[Kprawn]

I like to think this is rather something like this.. https://www.youtube.com/watch?v=ZWSMoE3A5DI  {Skip to 3:11} ... The people doing these tests know it's harmful but

they simply do not care, because the ultimate goal is to destroy the network. Do some small attacks and see how successful they are and then launch the final attack to sink

it, at a reduced cost. This is nothing other than sabotage, in a time when it's not welcomed. They do not care, if a few soldiers die in the process.     

[countryfree]

A typically reaction is to wait and combine several small transactions into a large one and send it with a higher fee (just like ATM cash withdraw in foreign country, you always do the maximum withdraw to reduce the impact of the fee since fee is fixed regardless of withdraw amount)

That depends on the bank. Usually, there are 2 fees. One from the bank which issued the credit card, and another from the foreign which provide the cash. One can be fixed, the other can be %, there are many cases...

Regarding the stress tests, I welcome them. They show what it will be like next year, with more people using BTC. Banks also do stress tests!

Stress tests don't have to be disruptive. That's why we have testnet.

This isn't a 'stress test', it's a premeditated attack on the network.

Good. I want more attacks. Especially from friendly BTC supporters, because we have to get ready in case of some real bad guys (Islamic State, Russia...) organizing a real attack to destroy western economies.

[MicroGuy]

Good. I want more attacks. Especially from friendly BTC supporters, because we have to get ready in case of some real bad guys (Islamic State, Russia...) organizing a real attack to destroy western economies.

True. But friendly bitcoin supporters would never do things that harm the network and disrupt the entire economy. Everyone already knows that the 1MB block size needs to be increased, so this is just an attack exploiting an already known vulnerability.

Finding a new weakness is one thing, but this attack is not helping bitcoin. It's unnecessary and counterproductive.

[tl121]

Good. I want more attacks. Especially from friendly BTC supporters, because we have to get ready in case of some real bad guys (Islamic State, Russia...) organizing a real attack to destroy western economies.

True. But friendly bitcoin supporters would never do things that harm the network and disrupt the entire economy. Everyone already knows that the 1MB block size needs to be increased, so this is just an attack exploiting an already known vulnerability.

Finding a new weakness is one thing, but this attack is not helping bitcoin. It's unnecessary and counterproductive.

The recent tests/demonstrations/attacks have already shown weaknesses in various bitcoin implementations.  For example, they have shown that the existing Electrum server code will be unable to keep up with higher network throughput possible once the block size has been expanded, even on very fast processors.  As a result, there is (finally) serious effort to upgrade this software.  I'm sure there are other examples, but this is the only one that I am personally familiar with.

[johnyj]

Good. I want more attacks. Especially from friendly BTC supporters, because we have to get ready in case of some real bad guys (Islamic State, Russia...) organizing a real attack to destroy western economies.

True. But friendly bitcoin supporters would never do things that harm the network and disrupt the entire economy. Everyone already knows that the 1MB block size needs to be increased, so this is just an attack exploiting an already known vulnerability.

Finding a new weakness is one thing, but this attack is not helping bitcoin. It's unnecessary and counterproductive.

The recent tests/demonstrations/attacks have already shown weaknesses in various bitcoin implementations.  For example, they have shown that the existing Electrum server code will be unable to keep up with higher network throughput possible once the block size has been expanded, even on very fast processors.  As a result, there is (finally) serious effort to upgrade this software.  I'm sure there are other examples, but this is the only one that I am personally familiar with.

That's good, just like the bank stress test, some of the banks will fail the test and have to improve their liquidity, if we could periodically have this kind of test, then we can have a good estimate of each node's capability in case a sudden surge in transaction volume hit (non-spaming, during price rally or crash)

[tmltd]

That's good, just like the bank stress test, some of the banks will fail the test and have to improve their liquidity, if we could periodically have this kind of test, then we can have a good estimate of each node's capability in case a sudden surge in transaction volume hit (non-spaming, during price rally or crash)

Don't compare a bank stress test with the current Bitcoin stress test.
During a bank stress test the customers of the bank will not notice any difference in their interaction with the bank. They don't pay more tx fees, they don't find longer queues in cashiers or ATMs. They don't even know that the bank is under an ongoing stress test. And finally the definition of bank stress test is: "An analysis or simulation designed to determine the ability of a given financial instrument or financial institution to deal with an economic crisis". The key words are analysis or simulation, not destructive activities that can block bank customers to interact with their accounts.
The current Bitcoin stress test is not an analysis or simulation using the testnet as it ought to be and finally will not produce any new results that we don't know yet.

[rico666]

And let me clarify upfront: The reasons for my positive attitude can be classified as "Techno-Darwinism".

Bitcoin and its underlying blockchain is an incredible piece of technology. The reason I write "piece" and not "masterpiece", is because there are still dark places. Some we know and I believe some we still don't know. From time to time these dark places appear on our radar and sometimes they stab us in the back unannounced.

Remember - among others - the "faucet" bitcoin client in 2010 that allowed you to create large amounts of bitcoin? Or the March 2013 hard fork?

I'm talking solely about the technical blows Bitcoin experienced, not the socio-economical like thefts, Go[of]xs and the like. Guess what happened *every single time*? That's right: What doesn't kill you makes you stronger. It may seem twisted to you, but following this, we should actually try to kill Bitcoin more often and in a more systematic way.

I do not propose Bitcoin XT. In fact, I hope it will vanish deep in the anals of history - so to speak. ;-)

On the other hand, the limits regarding transactions is a serious issue and anybody who would like Bitcoin to become more popular and widely accepted should agree that this problem (and a problem it is) has to be solved and it has to be solved for good.

For me personally, just increasing the max. block size is not a good solution, but then again I have not contributed a single line to the core/client. Some adaptive mechanism would be better, and a more generic concept of offloading transactions (merchant sidechains) probably even more so.

If nothing else, this stress test would lead us somewhere where we have never been before. Reminds me of a recent session with https://loader.io/ where we tested a web site. The productive one, I might add. Result: About 2 hours of intermittent downtime and bad availability, but in the end the site is now able to flawlessly cope with twice as many concurrent clients (1000) than before when 500 started to take it down. Plus we now have better insight where and how to scale.

If someone is willing to spend 100 BTC to enable(!) the Bitcoin community to do a similar thing and announces this in advance, I cheer to him. Bitcoin is still very young and still has to prove itself. Events like these contribute to the proof of concept.


Rico

[ranochigo]

Satoshi-dice was sort of the stress-test of it's day.  It was annoying, but ultimately I think it was one of the best things that happened for the system since it spurred a range of developments (code and otherwise.)  I said so at the time IIRC.  Overcoming that challenge gave me and I suspect other people more confidence in the system.

Spam from Satoshi-dice and other onchain gambling sites are easy to filter. Miners can easily change their code such that the transactions orginating or   going to their known address would effectively not be included in the block. However, the main problem with such stress test attacks is that the coins can be sent from different address and end up in different address. Without a huge amount of fees per transaction, they cannot cripple the entire system and it would be quite expensive. Currently, many attacks are just slowing down the confirmation timing for transactions with very low fees. The most viable way of stopping this kind of spam is to increase block size, thus increasing the cost for attackers to fill the block with spams.