In short, there was some good and some not as good. The good news is most of the ‘bad’ things are related to known throughput limitations of the current iteration of Stacks. The congestion issues and the symptoms they cause were certainly present as the coordinated nature of the release created an all-at-once influx of transactions.
Further, exchanges aren’t immune to impacts from surges like the one on Sunday. Two exchanges unprepared for the large uptick in activity without automated fee tools experienced long delays due to transactions with low fees. Obviously, this also impacted users who were likely caught off guard by the need for much higher fees than normal. (If your tx is stuck, remember
you can increase the fee!).
On the positive side, other integrations such as bridges, oracles, custody providers, etc. seem to be unaffected sans adjusting fees where needed. Further, the Stacks layer was able to keep chugging through transactions, even compiling one block with over 8,000 transactions and another of over 10,000, a new record for the layer. For reference, the average number of transactions per day in November was 3,865, but since Friday evening that average is up to ~32,000 per day (total of approximately 126,000).
Recent improvements to mempool handling and other optimizations helped the Stacks layer deal with the spike, albeit still more slowly than we’d all like. Notably, thanks to recent optimizations to the transaction selection process, we saw miners able to build a full block in 10 seconds, whereas before, they couldn’t even fill it 2.5% in a full 30-second window. That’s 3 times faster for 40 times the block size.