Question: Is this project still maintained?

[chobostar]

No releases since december 2024

Here is information from the maintainer (@hughcapet) about Spilo aliveness: https://github.com/zalando/spilo/issues/1131#issuecomment-2956743312

Since the Postgres Operator depends on Spilo, should we expect this project to be dead as well?

Thanks for the clarification in advance!

[FxKu]

This project is indeed in a little idle state at the moment. After pushing for more frequent releases last year I had to focus on some bigger internal projects at Zalando. We have a tracking sheet for PRs we aim to include in the next release which is probably gonna happen in July.

Fact is, we still use Spilo and the Postgres Operator in production, so we will keep maintaining them to e.g. support new Postgres versions and probably also include newer K8s feature if it's required for us. Are you already using this operator or just evaluating?

It's clear that both projects show their age, but modernizing them was not encouraged by the management because "it works". Spending time/money on it would be challenged with switching to other cloud offerings instead. Sure, for most of the 1000s of database clusters my team manages our stack still works great and smoothly.

The most important asset is Patroni anyway, and this project is moving forward fast in comparison. You can be sure that breaking changes in Patroni will be reflected with care in the operator and in Spilo because we have a core contributor in our team with @hughcapet.

I hope to find more time again for the operator in autumn. For Spilo, though, I would not hope for much change.

[chobostar]

Thanks for the answer.

Are you already using this operator or just evaluating?

Yes, we have been using the operator for a couple years for now in production.

There are little inconviences regarding standby_cluster support that we use for DR solutions. I had considered contributing a PR to either the operator or Spilo, but I noticed that there are a lot of PRs that seem to be inactive or waiting on feedback.

[emrahbecer]

This project is indeed in a little idle state at the moment. After pushing for more frequent releases last year I had to focus on some bigger internal projects at Zalando. We have a tracking sheet for PRs we aim to include in the next release which is probably gonna happen in July.

Fact is, we still use Spilo and the Postgres Operator in production, so we will keep maintaining them to e.g. support new Postgres versions and probably also include newer K8s feature if it's required for us. Are you already using this operator or just evaluating?

It's clear that both projects show their age, but modernizing them was not encouraged by the management because "it works". Spending time/money on it would be challenged with switching to other cloud offerings instead. Sure, for most of the 1000s of database clusters my team manages our stack still works great and smoothly.

The most important asset is Patroni anyway, and this project is moving forward fast in comparison. You can be sure that breaking changes in Patroni will be reflected with care in the operator and in Spilo because we have a core contributor in our team with @hughcapet.

I hope to find more time again for the operator in autumn. For Spilo, though, I would not hope for much change.

Hi @FxKu ,

Thank you for the clarification about the status of the dependent projects. We've been using Zalando Postgres Operator in production for 4 years and we're happy with it. We appreciate your efforts.

It looks like the zalando postgres operator project has started losing momentum. Do you think production users should may be start considering alternatives like CloudNativePG which is more popular among K8S community these days? Can you please advice?

We've built production systems that depend on Zalando Postgres operator and spilo however if the future of these projects is not much clear we think may be we should have a back out strategy.

Thank you. Emrah.

[mortenlj]

We are in the process of setting up Zalando Postgres Operator in our environments for use in production, planning to replace Google CloudSQL databases (estimated cost reduction of around 50% in our case). We tested CloudNativePG and Zalando and decided on Zalando because it seemed more stable and robust.

I can understand company priorities holding you back from doing major changes, but what are your thoughts on PRs from external contributors?

We might be willing to invest in PRs to improve areas that are important to us, but that requires there to be capacity for reviewing, approving and merging PRs, and making new releases.

[robermar23]

We also use Zalando Postgres Operator in our environments for use in production and have come to love and depend on its functionality. We would love to contribute as well to enhancing standby_cluster support if PR's were reviewed and merged.