Cloud run is what i expected Appengine to be - A container plantform without any proprietary API’s or restrictions. Cloudrun recently announced GA.

Pro’s

Scale to zero container instances when there is no traffic.
Built on Knative and its portable across platforms and minimize lock-in
Pay-per use model - pay only for compute/mem/network used.
If you have existing GKE cluster then there is an add-on to use CloudRun on the existing GKE cluster.

Con’s

Currently you cannot check the current number of active instances. It isn’t clear if the instance was shutdown when there is no traffic. Additional reference: link here
Cannot configure CPU on fully-managed instance. Its possible only with on-prem Anthos.

While building from docker - files need to be in current directory of host and not target subdirectory.

  Step 2/4 : COPY HelloStackExplorer-0.0.1-SNAPSHOT.jar  /HelloStackExplorer-0.0.1-SNAPSHOT.jar
  COPY failed: stat /var/lib/docker/tmp/docker-builder188329130/HelloStackExplorer-0.0.1-SNAPSHOT.jar: no such file or directory
  ERROR
  ERROR: build step 0 "gcr.io/cloud-builders/docker" failed: exit status 1
  --------------------------------------------------------------------------------

  ERROR: (gcloud.builds.submit) build 7a43b056-0ac3-4fba-bed6-33b59fcff9b1 completed with status "FAILURE"

CloudRun example

In order to use CloudRun you just need to buld and deploy your application as container with any language/tools of your choice. Following steps give an example of running hello world web application:

Build the app
- Write code
- Define Docker file
As a quick hello-world example of app with docker file created the following: HelloWorldSpringBoot repo Just checkout the code locally for this step.
Publish Docker image to repo
```
gcloud builds submit --tag gcr.io/stackexplorer-project/hello
```
Published container is submitted to the private continer registery for your project. Its not visible to others.
Deploy the docker file to CloudRun
```
gcloud run deploy --tag gcr.io/stackexplorer-project/hello
```
Provide service name and region, after successful deployment it shows the

Alternatively you can deploy it to CloudRun on Existing GKE cluster using:

gcloud run deploy --tag gcr.io/stackexplorer-project/hello \
        --cluster=mygkecluster

Cost

You can manage costs by setting limit for memory allocated per instance, max/min number of instances. Reference: https://cloud.google.com/run/docs/configuring/max-instances

Example command to set memory limit and max instances:

 gcloud run deploy --image gcr.io/stackexplorer-project/hello \
            --max-instances 3 --memory 250  --platform managed