Last night I encountered that kind of "on-chain data suddenly stalls," and my mindset immediately collapsed for half a second... Later I thought it's probably not the chain stopping, but that the middle stuff is just gasping for air. Many front-ends actually ask the indexer/subgraph first (organizing messy on-chain transactions into queryable tables), and if it hasn't finished syncing, you'll see balances/TVL like they've been paused. Not to mention RPC rate limiting—if you refresh too fast, you'll get a 429, and the page can only spin in circles and pretend to be dead.

This lone wolf who watches data at night is very easily misled by this, thinking someone is dumping or the pool is being drained... but it’s just "data pipeline congestion." By the way, I’m a bit slow on that set of testnet points, everyone in the group is guessing whether the mainnet will issue tokens, and I’m still checking if the subgraph block height matches. Never mind, don’t let emotions take over; data delays are just delays.