The Ripple hardfork that Jed McCaleb and Joyce Kim initially released in July 2014, Stellar went down at the end of last week but the thing is, no one really cared?
The founder of Post Oak Labs, Tim Swanson first brought up the network being down last week in a tweet (see below) and a representative from Stellar later confirmed its occurrence.
Breaking: yesterday the Stellar network went down for about 2 hours… only those who run validators noticed it.
no new transactions were added for ~2 hours.
— Tim Swanson (@ofnumbers) May 16, 2019
Very few people noticed it and it kind of goes to show how little people care about the network. I mean we didn’t even hear about this until a week after it had happened!
Bear with me here, I’m aware this sounds like we’re digging into Stellar but the incident does raise a few questions about the level of decentralisation on the Stellar network. This uses the same ‘validator’ network design as Ripple does. In this design, only a few servers validate the transactions of the network at large. But if these servers do experience a rough patch then the whole system can potentially become unstable.
The chief scientist at Stellar has confirmed decentralisation is a worry for the protocol has he said that last month.
In the blog post, the scientist goes to respond to a separate report by researchers who conclude that Stellar is too centralised to be considered ‘secure’.
The researchers had said:
“As can be seen from many articles and papers, some network attacks, such as DDoS, can occur in the blockchain networks. In this paper, we are saying that if two centralized nodes can not receive or send any message because of DDoS, then all nodes in Stellar network wull [sic] be blocked and can not move to the next step in the consensus process.”
McCaleb has said that the Stellar Development Foundation nodes aren’t the only thing to blame though for the outage on the network. But he did say that the project has been working on getting people to rely less on these nodes.
“Over the last months we have worked to get people to *not* depend on the SDF nodes. As of maybe a month or so ago the SDF nodes could safely go down and the network would continue. But this also means that the network can halt even if the SDF nodes are still running. Unfortunately this is what happened. Enough other nodes stopped for various reasons that the network halted. The SDF nodes and in fact the majority of validators in the network were still up. They just couldn’t close ledgers safely because they weren’t hearing from enough nodes in their quorums so the network halted until it could be restored to a good state.”
Swanson points out that, were this to happen on another network, the effect would be more than doubled due to the fact that they are simply more active. Compared to other networks, Stellar is quite small in reality.
“What basically happened was that a critical mass of nodes went down causing a cascading failure and so the entire network went down but because it isn’t frequently used, few noticed.”
So when the network did go into ‘outage mode’, not many people noticed. It is the equivalent of a majority of miners going offline in Bitcoin and the remaining miners have effectively been unable to find blocks to give support for the network.
It was recently announced by Stellar that they would be teaming up with the guys at IBM. this boosted sentiments towards the cryptocurrency with Ripple and Stellar working in an increasingly fractured marketplace, where some banks have chosen to crate blockchain solutions as opposed to utilising these ‘industrial’ products.
If you are new to the space then your attention will probably be initially pointed towards Bitcoin or Ethereum but Stellar (XLM) is a project that is taking a different approach to the model of what a cryptocurrency is.
We are not by any extent ripping into Stellar here, it is still a project that is making waves throughout the crypto space, there’s no denying that. Although, no matter how you look at it, XLM isn’t as popular as XRP, BTC or even LTC.