New to Busy?

A Steem stress test experiment

38 comments

tarazkp
78
5 months agoSteemit4 min read

The highest transactions of any Blockchain by far is EOS and it has a record 24 hour period of 87 million transactions which is pretty impressive. According to https://blocktivity.info/, it is currently pushing about 23 million transactions a day and looks to be 17% utilized while Steem is pushing 1 million a day and is 0.1% utilized. So theoretically, if Steem was doing the same number of transaction, it would be at ~2% capacity.

Why don't we test how well the Steem Blockchain stacks up against EOS?

I don't know about the numbers here so someone with more skills than me needs to calculate this but:

An account with enough Resource Credits should be able to transact every 3 seconds for a grand total of 28,000 transaction in a 24 hour period. This means that if using this number, in order to equal EOS' record 3020 bots running at a maximum transaction speed would be needed. That is not so many but we could essentially double it to 6000 just in case to produce at a max 172,800,000 transactions and that would mean that if the Blocktivity numbers are accurate, Steem will have 17.2% utilization - The same as EOS transacting 23 million times.

Getting 6000 accounts (not mains) wouldn't be too hard I would predict and then they would each require enough delegation to allow them the RCs to transfer 28,800 times. A transfer is the cheapest of the transaction (other than votes) and each would cost 0.17M RCs. That won't require too much degation for each and there would be plenty of people with enough stake likely willing to delegate a bit to some of them for 24 hours for a trial.

One of the decent devs like @therealwolf could develop a site where alts could be staged, powered up with delegation and given a few Steem each that they would use to send and receive on a cycle in and out 1 > 2 > 3 and so on in a loop. The send, receive, send code shouldn't be too difficult to write I think but I do not know the intricacies.

While this would be a fantastic bit of publicity to see just what the blockchain is capable of (or an unmitigated disaster), the data it gives back to the development teams concerning real world tests for scalability could be invaluable for future innovations. The benefit of doing it this way rather than letting it reach it organically is that it can be tested in a somewhat controlled fashion and metered and dissected with precision.

As @abh12345 just commented:

Yes, this would be cheating, but it is a stress test that is inevitable and being able to engineer it early and then adjust if it all falls down means the least amount of people are affected - because if it fails when there are this many organic transactions it would mean there are likely a few million people here and then - that will get publicity.

And again as Asher said,

It would be interesting to watch.

 
Again, this would need more thought put into it than my math allows but if the steem blockchain is capable, it would definitely get some attention the cryptospace. According to Blocktivity, if looking at the percentage of utilization of the blockchain, only Steem and BTS are in the same class and if at 172 million transactions Steem could be at the same utilization as EOS with 23 million and blow WAX and TRX out of the water, it would make for some interesting press to read for once.

This might be a good way to test the blockchain after all the changes that Steemit Inc has implemented are in place as the volume would provide a mountain of leveragable performance data for all concerned. Of course, witnesses would have to get involved to see if this is viable too or whether it is too risky to test at this time as most people on the chain don't have the skills or knowledge to predict the outcome - especially me.

Anyone game for a Steem stress test?

Taraz
[ a Steem original ]

Comments

Sort byBest