isolator VS Sidekiq

In fact, the project had already taken advantage of a number of Evil Martians OSS products, like Autoprefixer, after_commit_everywhere, and isolator.

The first is isolator, a gem by the folks over at Evil Martians that detects non-atomic interactions within a database transaction. What do we mean by that? A simple example taken from the documentation that uses background jobs is something like the following:

add isolator

I'd say Sidekiq is the top competitor here.

There's something wrong at Redislabs, it took them over a year to get RESP3 rolled out into their hosted service, you'd expect a rollout of that to be a bit quicker when they're the owner of Redis.

It affected us when upgrading Sidekiq to version 7, which dropped support for older Redis, and their Envoy proxy setup didn't support HELLO and RESP3: https://github.com/sidekiq/sidekiq/issues/5594

That depends on how the `maxmemory-policy` is configured, and queue systems based on Redis will tell you not to allow eviction. https://github.com/sidekiq/sidekiq/wiki/Using-Redis#memory (it even logs a warnings if it detects your Redis is misconfigured IIRC).

Sidekiq https://sidekiq.org/: This one started as an open source project, once it got enough traction, the developer made a premium version of it, and makes money by selling licenses to businesses.

Sidekiq will drop in-progress jobs when a worker crashes. Sidekiq Pro can recover those jobs but with a large delay. Sidekiq is excellent overall but it’s not suitable for processing critical jobs with a low latency guarantee.

https://github.com/sidekiq/sidekiq/wiki/Reliability

> I'm not sure feature withholding has traditionally worked out well in the developer space.

I think it's worked out well for Sidekiq (https://sidekiq.org). I really like their model of layering valuable features between the OSS / Pro / Enterprise licenses.

I was studying Sidekiq's page on rate limiters. The first type of rate limiting mentioned is the concurrent limiter: only n tasks are allowed to run at any point in time. Note that this is independent of time units (e.g. per second), or how long they take to run. The only limitation is the number of concurrent tasks/requests.

The code above isn't idempotent. If you run it twice, it will create two copies, which is probably not what you intended. Why is this important? Because most backend job processors like Sidekiq don't make any guarantees that your jobs will run exactly once.