Trying to understand if your per-log throughput will be capped by the max traffic that can be pushed through a single sequencer, Yes. You can think of an indi

Trying to understand if your per-log throughput will be capped by the max traffic that can be pushed through a single sequencer, Yes. You can think of an individual log like a database shard. We have plans to allow the sequencer to just give out sequence numbers, and allow the clients to send the data directly to the storage nodes. But its not a high priority for us, since it's just a constant factor improvement. (Although a large one!) Users would still need to shard their data, although not as much.