Lately, I keep feeling like the "system is lagging" when looking at on-chain data, but actually, it's often not the chain itself that's slow, it's the layer you're viewing that is struggling: indexers need to scan logs, build indexes, subgraphs have to wait for new blocks to be processed, and when RPCs are rate-limited or queued, your frontend just spins as if it's offline... To put it simply, what we usually see as "real-time" is actually just cache + queue hard-trying.

It's a bit like the current disputes over NFT royalties: creators want stable income, while the market desires smoother liquidity, but ultimately, it comes down to how the infrastructure makes trade-offs and who takes the blame as the middle layer. Anyway, now I tend to open two RPC providers and compare the latency of different indexers to keep a calmer mindset... As for how you usually judge whether the chain is slow or the data layer is lagging, I’m still figuring that out.