Developers from Anza, Firedancer, Jito, and other core contributors are working diligently (and not sleeping much) to shore up Solana's networking stack to meet the unprecedented demand the network is seeing today.
There's been a lot of threads on what exactly is causing the extreme congestion – and many of those threads disagree with one another – but at a high level the issue is conceptually simple: the implementation of a software system is today not robust enough to handle the amount of traffic being thrown at it. That's now being fixed – but an obvious question is "How did we get here?"
This was, to some extent, a known bottleneck in the implementation of QUIC used on the Agave validator client, and was on the roadmap to be addressed. What was not expected was the rate at which demand for Solana would increase, flipping the implantation from 'adequate but needing improvement' to 'inadequate' in a matter of days. A charitable reading is a 'failure of success' which is otherwise known a 'failure of planing'
This is, put simply, tech debt. The tradeoff between new and important new work (like readying Solana to support its first independent validator client implementation, Firedancer, which includes near total rewrite of the networking stack, later this year) and important maintenance/improvement work. Every decision is a series of tradeoffs, sometimes you get it right, sometimes you get it wrong. This is not dissimilar from what the Solana network went through in early 2022 – when demand outstripped the capacity of several systems.
Core protocol developers from across the ecosystem are working as swiftly as they can to implement and test improvements in the networking stack to address the current network congestion. It will take some time, but we'll get there. This open source software project is nothing without the thousands of developers and millions of users who believe in the vision of Solana. We'll get there, together 🤝