MetaStreme stress test ‘clear evidence’ Bitcoin network wasn’t stressed at all

Last week, MetaStreme put the Bitcoin network to a test—a stress test that is. The new high throughput wallet service from the WeatherSV team broadcasted close to 1 million transactions on the Bitcoin SV network in 30 minutes and in the process, achieved over 600 transactions per second throughput rate.

The mainnet stress test had a two-fold goal, according to its founder Paul Chiari. The team needed to ensure that MetaStreme works as intended, while also gauging the computation costs of running a service.

The stress test, however, had another coup. As nChain CTO Steve Shadders pointed out:

The network ate this up. What we know about network behaviour is that it handles intense bursts of traffic very nicely. The real challenge is sustained throughput which we are rapidly improving on. https://t.co/FiAwtveNfE

— Shadders (@shadders333) June 24, 2020

In an interview with CoinGeek, Shadders explained what he meant when he said “the real challenge is sustained throughput.” He told us:

It is common for any capacity bound system to be able to handle short term peaks of inputs at much higher levels than they can sustain on a long term basis. As an analogy, consider touching a hot pan. If you touch it for just a second you can tolerate a much higher temperature input on your skin than if you held onto it for a full minute. Although this test was below what we believe the sustained capacity of Bitcoin SV’s mainnet to be, the fact that it was a short, controlled burst made it even easier for the network to handle and the results show that the nodes barely even noticed the spike.

Shadders said the MetaStreme stress test “wasn’t a test of the Bitcoin network capacity,” but was actually a test of the wallet application itself. However, it’s worth noting that the Bitcoin network handled the high level of traffic “without breaking a sweat.”

“[The stress test] is more a signal for potential users of MetaStreme that it is capable of handling the demonstrated throughput. It is also clear evidence that at least this level of traffic can be handled by the Bitcoin network without breaking a sweat,” Shadders said. “What we can simulate on the Scaling test network (STN) tells us a lot, but the production network is always going to be a more complex structure with different behavioral characteristics, so it is good to have the capacity test confirm lower bounds when they are done under controlled conditions like this was.”

Lessons learned from MetaStreme stress test

Chiari said they experienced network issues during the stress test, which highlighted “the importance of real world testing under load” to help the MetaStreme team come up with preventive measures especially since they are planning to open the platform to beta testers in the near future.

For Shadders, one of the key points about the successful stress test was that the Bitcoin SV Node team was notified shortly before it began, which he said allowed them to monitor the network state throughput allowing them “to gain insights on how the production network behave under loaded conditions.”

So for platforms and projects that are planning to conduct a mainnet test on BSV, Shadders has this piece of advice: communicate.

“First and foremost that communicating with the SV team before these sort of tests is not only helpful to the party conducting the test because we can offer advice, but helpful to Bitcoin SV as a whole because it gives us an opportunity to monitor and gather data during the event. The more notice the better because we need to make sure we have tools and people on standby for these situations.”