We noticed that the relayer cannot submit messages as fast as new ones arrive. This increases the latency of message delivery, affecting user experience and causing many false alarms in our monitoring.
After considering both pipelining and batching of message submission, we'll initially go with batching, as it can be shipped incrementally (initially EVM only, for chains that already support multicall).
Full context here: https://www.notion.so/hyperlanexyz/Improving-Relayer-Throughput-43462a3f52d547458e55d9c66ac93e2f
We noticed that the relayer cannot submit messages as fast as new ones arrive. This increases the latency of message delivery, affecting user experience and causing many false alarms in our monitoring.
After considering both pipelining and batching of message submission, we'll initially go with batching, as it can be shipped incrementally (initially EVM only, for chains that already support multicall).