plumber VS sqlite

Streamdal

4. Go to the provided UI (or run the CLI app) and be able to peek into what your app is reading or writing, like with `tail -f`.

And that's basically it. There's a bunch more functionality in the project but we find this to be the most immediately useful part. Every developer we've shown this to has said "I wish I had this at my gig at $company" - and we feel exactly the same. We are devs and this is what we’ve always wanted, hundreds of times - a way to just quickly look at the data our software is producing in real-time, without having to jump through any hoops.

If you want to learn more about the "why" and the origin of this project - you can read about it here: https://streamdal.com/manifesto

— — —

HOW DOES IT WORK?

The SDK establishes a long-running session with the server (using gRPC) and "listens" for commands that are forwarded to it all the way from the UI -> server -> SDK.

The commands are things like: "show me the data that you are currently consuming", "apply these rules to all data that you produce", "inspect the schema for all data", and so on.

The SDK interprets the command and either executes Wasm-based rules against the data it's processing or if it's a `tail` request - it'll send the data to the server, which will forward it to the UI for display.

The SDK IS part of the critical path but it does not have a dependency on the server. If the server is gone, you won't be able to use the UI or send commands to the SDKs, but that's about it - the SDKs will continue to work and attempt to reconnect to the server behind the scenes.

— — —

TECHNICAL BITS

The project consists of a lot of "buzzwordy" tech: we use gRPC, grpc-Web, protobuf, redis, Wasm, Deno, ReactFlow, and probably a few other things.

The server is written in Go, all of the Wasm is Rust and the UI is Typescript. There are SDKs for Go, Python, and Node. We chose these languages for the SDKs because we've been working in them daily for the past 10+ years.

The reasons for the tech choices are explained in detail here: https://docs.streamdal.com/en/resources-support/open-source/

— — —

LAST PART

OK, that's it. What do you think? Is it useful? Can we answer anything?

- If you like what you're seeing, give our repo a star: https://github.com/streamdal/streamdal

Checkout plumber https://github.com/streamdal/plumber if you want all the Postgres changes sent to basically any type of broker queue https://docs.streamdal.com/en/data-ingestion/relay/postgresql-cdc/. I would say Nat's Jetstream is probably the way to go if you have K8s running already. It's a dead simple service written in Go. Just make sure you allocate enough memory to Jetstream.

I recommend Streamdal. The connecting agent is open source and distributed by default, so it will scale horizontally WAY better than Debezium. All data ingested is indexed into parquet as well, and you can do serverless functions/transforms on the platform to reduce Snowflake compute costs.

For CDC and event driven part of the stack, Plumber is a great free tool. That project is going to be adding sampling soon too - this can def help with the cost of ETL.

https://github.com/batchcorp/plumber check it out if you want to get into event driven systems

That’s awesome to see. My coworker and I always figured Go would be perfect for this. Going to be a serious amount of work! I see you use NATS as well. Big fan of it. Checkout our project https://github.com/batchcorp/plumber if you end up needing to inspect or send messages while deving against it.

Checkout plumber https://github.com/batchcorp/plumber join our slack https://launchpass.com/streamdal we got a pretty knowledgeable group

> I would've probably picked the modernc variation

Heads up about the modernc library, it has been stuck on an old version of sqlite for several months [1]. It seems like maintainer time is the limiting factor [2]. There has been a call to arms on that issue page, the maintainer is looking for help, but it looks like not much has arrived. It seems like it might trace back to blockers in the C-to-Go compiler.

It's a major undertaking and a very impressive piece of work, but I'm not surprised it's a struggle when big roadblocks get hit. I hope they find a way to progress, but I'm very relieved to be seeing some CGo-free alternatives like ncruces/go-sqlite3 emerging. I'm going to give it a try for sure and see if I can live with the compromises.

Squinn-go looks very compelling too, but I don't like that it requires the squinn binary to already be installed on a user's machine, I think that gives with one hand and takes with the other: sure, I get to avoid CGo, but I also lose the turnkey, single-command install, static build benefits Go brings out of the box.

Seconding the point about nitty gritty, I'd read it for sure too!

  [1]: https://gitlab.com/cznic/sqlite/-/issues/154

No, but that has the disadvantage of being C compiled into Go, then being compiled into native executable.

I'm actually surprised by how readable this came out; props to the Go->C compiler author. But you can guess that pushing this sort of thing through the Go compiler is going to cause some slowdowns due to sheer paradigm mismatch: https://gitlab.com/cznic/sqlite/-/blob/master/lib/sqlite_lin...

FWIW, we use a version of SQLite transpiled into Go to avoid CGI problems: https://gitlab.com/cznic/sqlite

It could be clearer in the readme, but note that this is a machine translation from C to Go, repeated for every OS-Arch pair. Example of the one you're most likely to use in production: https://gitlab.com/cznic/sqlite/-/blob/master/lib/sqlite_linux_amd64.go