fix: rename Docker image

[ntn-x2]

The Docker image specified there never existed, and it's only used internally for cache.

[rflechtner]

The kilt/prototype-chain image very much did exist on our ECR and we're using it everywhere in our sdk tests.
@ggera can you confirm which image name we should be using? Can we maybe just configure ECS to act as a pull-through cache for kiltprotocol/standalone-node?

[ntn-x2]

Yes it does, but as far as I can see it's only used as an intermediate cache to build standalone-node Docker images. If @ggera agrees, we should probably switch SDK tests to use the standalone-node Docker image, and in the future perhaps even a Zombienet-based setup with the kilt-node Docker image directly, which I am working on.

[ntn-x2]

Anyway I'm a bit confused, cause here we use kiltprotocol/standalone-node:latest-develop instead.

[ntn-x2]

It seems we are not pushing there anymore @ggera? That is what I figured by looking at our deployment pipelines, although I think it was introduced with your refactoring. Since it's used as a caching image, perhaps we want to replace it completely in our build pipeline and in the integration tests to use the standalone image rather? This one has pushes from the current release process.

[ntn-x2]

Link to the manually-triggered GitHub workflow being changed here: https://github.com/KILTprotocol/sdk-js/actions/runs/10717708953/job/29718068093. This seems to work now.

[Dudleyneedham]

LGTM

[rflechtner]

We need a cache layer to get around dockerhub rate limits. That's why I suggested a pull through cache

[rflechtner]

Is standalone-node an image that's being updated on our ECR?

[rflechtner]

If this is the image we should be using, can you update all workflows to use it?

[ntn-x2]

Yes, standalone-node has always been updated, that's why I was confused. I still don't know if stop pushing to the prototype-chain image was intentional or not, but I can definitely update the image used everywhere.

[ntn-x2]

@rflechtner I talked to @ggera and we don't see any issue in using the final Docker image for all tests, so I updated all references. Once he's back, we'll investigate if we need such an intermediate image, and how to deal with it.

[rflechtner]

Lgtm

[ntn-x2]

The CI fails because the latest tag is broken, but both the latest rc as well as develop work, so I'll merge this. latest will be fixed as soon as we promote the rc.