How to put large-scale data?

I'm trying to store a large amount of data (tens of thousands or more) in an etcd cluster using jetcd. Currently, I'm struggling with how to efficiently perform multiple put operations in parallel. I understand that jetcd does not natively support batch operations, but the current approach of executing put one by one seems inefficient for large volumes of data. I’m concerned about the performance impact and the overhead of creating separate requests for each key-value pair.

As an alternative, I've considered manually executing multiple put operations concurrently using CompletableFuture to parallelize requests. However, this requires careful management of network resources, connection pooling, and error handling to avoid overloading the etcd cluster and potentially causing issues with rate limits or timeouts.

So I'd like to ask the official opinion on How to put large-scale data?