Migrate to async_nats NATS client

[makarchuk]

A note for the community

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Use Cases

Vector currently uses nats::asynk client which is not a recommended way of using it, as newer official async api exists in a form of async_nats crate. nats crate is described as "Legacy synchronous client that supports" in it's own repo, which also mentions that This client will be deprecated soon.

Furthermore nats::asynk lacks async support for NATS JetStream (there's a sync one, but that doesn't fit for vector's purposes), and JetStream seems to be a pretty good candidate for becoming another Vector source (I'll create a separate feature request for this later and will try to contribute this client myself).

So it seems that migrating to a more modern and async native async_nats client would be benifitial for a project as a whole.

Wanted to mention that I'm sorry if I failed to follow a proper process here. I tried my best and please feel free to point out any mistakes I've made.

Attempted Solutions

No response

Proposal

I have prepared an MR that implements such a migration. Although there're a couple of things that concern me and I would like to hear an opinion on those:

• async_nats doesn't support reconnect_buffer_size configuration option and judging by a comment in Vector code where this options is set Set reconnect_buffer_size on the nats client to 0 bytes so that the client doesn't buffer internally (to avoid message loss). it appears to be quite crucial. I can create a feature request in https://github.com/nats-io/nats.rs as it seems to me it's a matter of simple addition of a setter method.
• Due to changes in error handling approach nats client no longer returns std::io::Error as it's error type, so I was not able to retain error_type field in internal_events/nats.rs

References

#15533

Version

No response

[paolobarbolini]

FYI I'm making a new attempt at migrating to it in #18165