3 ms·
It’s already quite bad unfortunately. Both support for read replicas and metrics(!) are not in the open source version. They have a prometheus exporter but not
by klabb3 4y ago
It’s already quite bad unfortunately. Both support for read replicas and metrics(!) are not in the open source version. They have a prometheus exporter but not on free, afaik.
I was also concerned with people reporting memory consumption/leak issues, as I’m planning to have lots of subscriptions. I don’t know haskell well enough, but from the outside it does match the symptoms of having dug themselves into an architectural complexity hole, which I assume is much harder to navigate with unorthodox tooling and practices.
Im still rooting for hasura, I like their way of making things dumb, simple and without too much technophilia for its own sake. They genuinely want you to focus on your business problems instead of blasting you with novel and overengineered concepts, like others do.
- zwily 4y agoYeah, I love the product. I’ve been using it a few years in a fairly complicated internal tool used by a dozen people, and it’s been amazing. But I’m pretty discouraged by what I see as its future.
- tango12 4y ago(from Hasura) Scaling subscriptions is hard, but we work with our users/customers at scale to make sure settings are tweaked correctly. We have users running 100k - 1M concurrent users in production for live-event type platforms. It's not completely trivial to benchmark and setup because query patterns, streaming vs live queries etc have an impact, but it works very reliably. No missing events, no problems disconnecting/reconnecting, no need for sticky sessions and so on. An initial POC benchmark [1] should be a quick affair so if you're trying it out and run into any problems, please hit me up! Email on my bio. [1]: https://github.com/hasura/graphql-bench https://github.com/hasura/graphql-bench
- klabb3 4y agoThanks, that looks reassuring. I will let share my own benchmarks too if I get around to it.