Curriculum·J301 Chains, Layer 2s, and Bridges·about 32 min
Alt-L1s
By the end of this lesson you can
- →Identify what a chain gave up to obtain the throughput figure it advertises
- →Explain why uptime stated as a percentage understates what an outage costs
- →Compute the exposure a halt creates on a position you cannot close
- →Evaluate a chain by validator count, client diversity, halt history and restart authority
Junior · enrolled learners
This lesson opens with The Solana halt, 14 September 2021.
- What happened
- Solana's mainnet stalled following a token launch on Raydium's AcceleRaytor platform. Transaction load peaked at around 400,000 per second, driven by bots. One bot structured its transactions to write-lock eighteen key accounts, including the global SPL token program and the Serum DEX program, which blocked every transaction touching those accounts and destroyed the parallel execution the chain depends on. The queue flooded, network-critical messaging was not prioritized, and the network began to fork. The chain was down for roughly seventeen hours until validators manually coordinated a restart. Further outages followed through 2021 and 2022, and independent trackers maintain a full history.
- The decision point
- Seventeen hours is 0.19 percent of a year, so a chain with this outage still reports better than 99.8 percent uptime, and that figure is the wrong way to think about it. Outages happen under extreme load, extreme load arrives with volatility, and volatility is the only condition under which anybody urgently needs to act. The downtime is therefore not randomly distributed across the year. It is concentrated on the days when being unable to transact costs the most, which is the opposite of what an availability percentage implies.
What you will be able to answer
- →Why is uptime as a percentage misleading for a chain?
- →What did the Solana halt actually cost participants?
- →What does a chain usually trade for throughput?
- →What are the four evaluation questions?
Orientation and Year One are open: anyone can read them without an account. From Year Two onward the lessons are for enrolled learners, because progress through the later years only means anything if it is tracked against a record.
It is free. We do not sell the list and there is nothing to buy at the end of it.
Sources and review
- https://www.helius.dev/blog/solana-outages-complete-history
- https://www.coindesk.com/markets/2021/09/14/solana-validators-ready-potential-restart-amid-blockchain-outage
- https://www.theblock.co/amp/linked/117711/solana-blockchain-validators-restart-network-after-transaction-stoppage
- https://l2beat.com/scaling/summary
Confidence high·Volatility medium·Reviewed 2026-08-06·Owner unassigned
Contested
Naming Solana is not a claim that it is uniquely fragile or that its engineering has not changed since 2021. It is the best documented halt in the sector and the mechanism is instructive. Several chains have halted, and a chain that has never halted has either solved the problem or has not yet been tested at load, which are different things that look identical from outside.
Whether a coordinated restart is a defect or a safety property is genuinely contested. A chain that can be restarted by its validators can also be stopped by them, and this lesson presents that as a tradeoff rather than a verdict.