api design, should we use uuids or sequential ids for primary keys

Hey everyone, I'm trying to figure out the best way to handle primary keys for our new microservices. We've always used sequential integers, but now that we're going with distributed systems, and might need to merge data or shard later on, UUIDs are looking pretty good. On the flip side, sequential IDs are just simpler and often easier to debug. What do you all think? Have you used UUIDs versus sequential IDs in API design? I'm curious about the performance, database indexing, and just general ease of use.