telepresence | Local development against a remote Kubernetes | Continuous Deployment library

I have been working with Kubernetes for container orchestration and I have been using telepresence for swapping deployments from local to dev cluster and works like a charm. This also helps me in testing and fast iterations as I do not have to build and wait for my code to be deployed to the dev cluster.

Recently I switched to ECS and I have been wondering if there is a way to swap services from ECS Farget dev cluster to service running locally on my laptop.

I have this bash function in my .bash_profile:

I'm currently installing the Ambassador Edge Stack (AES) to help manage several applications running applications in our GKE cluster but I'm experiencing a couple of issues.

The steps in the manual install guide seem to be working fine aside from edgectl being deprecated in favour of telepresence (I haven't tried this out really).

The next steps, setting up the ingress with GKE is where the issues begin.

As per the guide, this can be done with the legacy Ambassador API Gateway or the new AES. Doing a compare of both installs, you don't need to do anything in the API Gateway install other than sorting out patching the original aes ambassador service and the ambassador-admin service from LoadBalancer to NodePort types. I've done this with kustomize

I have several pods where the rollout (initial and later updates) needs to performed one by one. (Actually only the first needs to be ready before the remaining can start or be upgraded) I used stateful set for that, as it makes sure only one at a time is updated or created, but we are usung telepresence for development and it doesn’t support replacing stateful sets. So I thought I could use deployment instead of stateful set with rolling update strategy and limit the number of maxunavailable or maxsurge or whatever to “throttle” the deployment. But for the initial deploymemt that doesn’t work, as K8s creates the desired 2 at once instead one one by one. Is there a way to achieve that with a deploment or do I need to use a stateful set? (alternatively: is there a trick to use telepresence with stateful sets)

Clarification based on questions in comments:

• The problematic software here is flyway in combination with mariadb in cluster mode. Then the table locking doesn't work and simultaniously starting pods can try to perform schema and data updates at the same time
• init containers don't help, as they start at the same time for multiple instances of the pod and just make sure that the main container of each instance is started after the init container
• The problem is only on first initalization, because afterwards I can configure the rolling update strategy to only update one container at a time. In case of a scaling out I'd have to do it in increments of 1, but that would be a manual process anyway.
• I could make sure that the deployment descriptor for new deployments uses scale 1 and updates to scale 2 afterwards, but that leads to a very complicated automatic deployment process with variable scales dependig on the state and the build chain would need to check if a deployment is present to decide if it's an update or a first deployment. Which would be error prone and overly complex

I'm used to connect to my cluster using telepresence and access cluster services locally.

Now, I need to make services in the cluster available to a group of applications that are running in docker containers locally. We can say that it's the inverse use case.

I've an app that is running in a docker container. It access services that are deploy using docker-compose. It has been done by using a network:

When I am using telepresence to debug remote kubernetes,throw this error:

when I start telepresence(v0.104) to debug remote kubernetes(v1.15.2) pod:

I'm makeing a table in Powershell like so :