nikita

24 ms·

271.
by nikita 4y ago
I think we need more details to answer this question. Can you describe your workflow with more details? Maybe even with some code examples ?
272.
by nikita 4y ago
AlloyDB is not serverless, cockroach is not Postgres, and fly doesn’t separate storage and compute.
273.
by nikita 4y ago
We would love to work with you on this. Please email beta@neon.tech.
274.
by nikita 4y ago
Systems >> databases for the engine. We also need fullstack or product engineers - modern DevX requires good UI. We also hire SREs and support
275.
by nikita 4y ago
Prob 60-70% margin over what is costs us.
276.
by nikita 4y ago
The answer is more pageservers. Right now we have 1 to may relationships 1 page server many tenants. But one tenant one pageserver. We will shard pageservers and make it many to many. The good news is that pageserver workload is constant sp
277.
by nikita 4y ago
We will onboard you sooner in return for honest feedback and working with us through the early wrinkles. Please email beta@neon.tech
278.
by nikita 4y ago
Yes, SQL Azure has customers with 10K+ tenants. I's anecdotal, but I'm pretty sure since I used to work at SQL Server! This makes it super easy.
279.
by nikita 4y ago
It's there!
280.
by nikita 4y ago
It's 3. Our intention is to only monetize DbaaS revenue and opensource all/most of the tech with Apache 2.0 license. It's similar to that of Databricks. Databricks over time built out photon that is proprietary. We will stay
281.
by nikita 4y ago
Yes. The issue is that compute doesn’t have enough of the cache AND you need to have a scratch space to update pages. You could theoretically do it on the compute too and this would be a valid design. A bit harder to work with read replicas
282.
by nikita 4y ago
Absolutely! We are working on it right now and call this “regions”. We already have a proxy - you will notice that the connection string is project_name.cloud.neon.tech. We are working on deploying the proxy globally and routing read traffi
283.
by nikita 4y ago
Thank you! Somehow open source technologies often don’t fully embrace cloud. This is an attempt to change that.
284.
by nikita 4y ago
This was a good discussion indeed. The blog captures most of these points. We were pre launch then
285.
by nikita 4y ago
People say that. The underlying tech is different. It’s closer to aurora than to Planetscale. Planetscale is shared nothing which breaks MySQL compat. We are 100% Postgres compatible.
286.
by nikita 4y ago
Only serverless, only Postgres - that’s focus. Multi-cloud and open source.
287.
by nikita 4y ago
Yes. There no dependencies on s3. Any object store will do. The next hill for us to take is more regions. Then more clouds
288.
by nikita 4y ago
Neon CEO here. Happy to AMA
289.
With Neondatabase, *truly* serverless PostgreSQL is finally here(twitter.com)
18 points by nikita 4y ago|4 comments
290.
by nikita 4y ago
Neon CEO here. Happy to answer any questions about Neon technology
291.
by nikita 4y ago
This is precisely right. Another point is that separation of storage and compute is decisively the right architecture for the cloud. Large cloud providers build their own implementation of this architecture which fragments the ecosystem. Ha
292.
by nikita 4y ago
We are exploring those things. You can do it both ways. Self host storage and let us manage compute. This way all data is in your infrastructure. We can also have you manage compute this way you can have more control over you customer. Righ
293.
by nikita 4y ago
Yes. Also if you run your app on AWS Lambda it's a lot easier to integrate with other serverless infra.
294.
by nikita 4y ago
The cloud version is at neon.tech. The foundation of serverless is separation of storage and compute. This cloud native storage is proprietary at AWS Aurora (and architecture is old). Neons storage is open source
295.
by nikita 4y ago
We are limiting the amount of compute we are giving for free during the technical preview. We will open it up to much large computes later. Like those will be in the paid tier.
296.
by nikita 4y ago
If it becomes more widespread and mature yes. We are in contact with Alexander Korotkov - Postgres committer and the auther of OrielDB and compare notes from time to time. Right now it makes a lot more sense to us to build our storage at th
297.
by nikita 4y ago
We will put more of the Aurora differentiation docs up in a bit. Architecture wise we are slightly different b/c of the S3 integration and splitting storage to safekeepers and page servers. This is very technical and we will be highlig
298.
by nikita 4y ago
We already have a shared slack channel with Kurt - CEO of fly and a number of folks from Fly.io. We are discussing our partnership.
299.
by nikita 4y ago
We will be launching REST api with PostgREST later this year.
300.
by nikita 4y ago
Major versions don't come out suddenly. We have time to test and make sure they work well with our storage. Customers don't always want to upgrade to new version when old versions just work. However this can lead to version creep.

More ›