Inquiry about Test 12

That’s Awesome :sunglasses:

6 Likes

It was partly a joke, but it was also meant to give a signal to people with slower connections that they won’t give up on them. So in this case you needed a 6 Mbit/sec. upload speed. But the team will probably work to give slower nodes a role as a Vault in the network’as well. Probably not routing data Chunks but think more of security or something.

TEST 12 went very well btw. If you check the interview with Viv he explains that it was around 4 months of work to get Disjoint Groups implemented. And read this about the security effect of that:

The group messages do not provide a high level of protection against attacks on the network, even under the additional assumptions that routes are disjoint and that a quorum of the individual messages need to be intercepted. Assume that an attacker controls 10% of all nodes in the network, and group size is 8 and quorum size is 5. Consider a section message sent via 10 hops.

In that scenario, the chance to intercept any given group request is 71%. With the section-to-section hops proposed here, it will be just 0.43%. Further adjusting the section size and quorum size parameters, this can be reduced to a negligible number, making this kind of attack virtually impossible.

So Routing became way more safe and secure. The possibility of an attack by a group that owns 10% of all Vaults got smashed into the bushes. Combine this with “Node ageing” and SAFE will be even more secure to attacks.

TL;DR : Four months on a very big refactor and only some small issues showed up :yum:.

Call me a fanboy (I am!) but what the devs are doing is amazing. Hopefully you’ll be able to connect next time as well. My Vault was connected in 5 minutes and stayed online for almost 24H.

13 Likes

From my understanding there were a few bugs as expected with ANY software. Even more so with something this cutting edge. Keep in my mind that the TEST 12 was DELIBERATELY shutdown due to the massive log file being generated by all of the network activity. Any longer and it would take the Maidsafe team weeks if not months to sort and analyze. That would make for slow development.

To remain agile they decided to pause things until they could ensure a higher level of stability. Few if any people tune a vehicle while in motion friend. Stop, go, and repeat until you get the desired behavior. The current community driven network proves that they could have left the TEST network running for likely a much longer period without catastrophic failure.

4 Likes

As far as I saw it, the bugs were to do with tunneling, plus a few minor issues, plus overly strict join policy. I had a vault running the whole time, serviced thousands of get requests, had a fast download for safe sites… could stream video realtime in a few tabs at once. From my vantage point, it was excellent. No need to keep it running and comb through gigs of log, instead of a few hundred megs.

9 Likes

For real. That test was amazing!! Latency and throughput were at reasonable levels considering the unoptimized nature of this test. Overall UX was nothing to laugh at. :sunglasses:

3 Likes

Were you using SAFEtube for that? Because all those videos are heavily compressed / downscaled. Just wondering if you were doing that or something like the tears of steel one, which would be very different

Compressed vids so not ridiculously exciting… the test was too short (and during work) to set up harder tests. Link to tears of steel? I uploaded some 720p stuff to try and test that but ran into the issue where is was tricky to setup a streamable video link from my test site. Pure D/L speed for the 720p videos was ~ 500 kByte/s after 6 hours on the network, easily enough to stream if I’d been able to work out how in time. Next time will be much better prepared for some bandwidth tests.

2 Likes