chore: Add Container Engine and Image Tag parameters to backend Makefiles

backend/Makefile

@@ -2,6 +2,15 @@ BUILD=build
 MOD_ROOT=..
 CSV_PATH=backend/third_party_licenses
 
+# Container Build Params
+CONTAINER_ENGINE ?= docker
+IMG_TAG_APISERVER         ?= apiserver
+IMG_TAG_PERSISTENCEAGENT  ?= persistence-agent
+IMG_TAG_CACHESERVER       ?= cache-server
+IMG_TAG_SCHEDULEDWORKFLOW ?= scheduledworkflow
+IMG_TAG_VIEWERCONTROLLER  ?= viewercontroller
+IMG_TAG_VISUALIZATION     ?= visualization
+
 # Whenever build command for any of the binaries change, we should update them both here and in backend/Dockerfiles.
 
 .PHONY: all
@@ -55,19 +64,19 @@ image_all: image_apiserver image_persistence_agent image_cache image_swf image_v
 
 .PHONY: image_apiserver
 image_apiserver:
-	cd $(MOD_ROOT) && docker build -t apiserver -f backend/Dockerfile .
+	cd $(MOD_ROOT) && ${CONTAINER_ENGINE} build -t ${IMG_TAG_APISERVER} -f backend/Dockerfile .
 .PHONY: image_persistence_agent
 image_persistence_agent:
-	cd $(MOD_ROOT) && docker build -t persistence-agent -f backend/Dockerfile.persistenceagent .
+	cd $(MOD_ROOT) && ${CONTAINER_ENGINE} build -t ${IMG_TAG_PERSISTENCEAGENT} -f backend/Dockerfile.persistenceagent .
 .PHONY: image_cache
 image_cache:
-	cd $(MOD_ROOT) && docker build -t cache-server -f backend/Dockerfile.cacheserver .
+	cd $(MOD_ROOT) && ${CONTAINER_ENGINE} build -t ${IMG_TAG_CACHESERVER} -f backend/Dockerfile.cacheserver .
 .PHONY: image_swf
 image_swf:
-	cd $(MOD_ROOT) && docker build -t scheduledworkflow -f backend/Dockerfile.scheduledworkflow .
+	cd $(MOD_ROOT) && ${CONTAINER_ENGINE} build -t ${IMG_TAG_SCHEDULEDWORKFLOW} -f backend/Dockerfile.scheduledworkflow .
 .PHONY: image_viewer
 image_viewer:
-	cd $(MOD_ROOT) && docker build -t viewercontroller -f backend/Dockerfile.viewercontroller .
+	cd $(MOD_ROOT) && ${CONTAINER_ENGINE} build -t ${IMG_TAG_VIEWERCONTROLLER} -f backend/Dockerfile.viewercontroller .
 .PHONY: image_visualization
 image_visualization:
-	cd $(MOD_ROOT) && docker build -t visualization -f backend/Dockerfile.visualization .
+	cd $(MOD_ROOT) && ${CONTAINER_ENGINE} build -t ${IMG_TAG_VISUALIZATION} -f backend/Dockerfile.visualization .

backend/api/Makefile

@@ -21,10 +21,12 @@ REMOTE_IMAGE=gcr.io/ml-pipeline-test/api-generator
 # Keep in sync with the version used in test/release/Dockerfile.release
 PREBUILT_REMOTE_IMAGE=gcr.io/ml-pipeline-test/api-generator@sha256:41fd3e60ba40430a4c3d87e03be817c5f63b2dfed23059ec9d6bca62ce0cc39c
 
+CONTAINER_ENGINE ?= docker
+
 # Generate clients using a pre-built api-generator image.
 .PHONY: generate
 generate: fetch-dependencies hack/generator.sh $(API_VERSION)/*.proto
-	docker run --interactive --rm \
+	${CONTAINER_ENGINE} run --interactive --rm \
 		-e API_VERSION=$(API_VERSION) \
 		--user $$(id -u):$$(id -g) \
 		--mount type=bind,source="$$(pwd)/../..",target=/go/src/github.com/kubeflow/pipelines \
@@ -44,7 +46,7 @@ v2beta1/google/rpc/status.proto:
 # we should push the new image remotely to ensure everyone is using the same tools.
 .PHONY: generate-from-scratch
 generate-from-scratch: .image-built hack/generator.sh $(API_VERSION)/*.proto
-	docker run --interactive --rm \
+	${CONTAINER_ENGINE} run --interactive --rm \
 		-e API_VERSION=$(API_VERSION) \
 	    --user $$(id -u):$$(id -g) \
 		--mount type=bind,source="$$(pwd)/../..",target=/go/src/github.com/kubeflow/pipelines \
@@ -57,10 +59,10 @@ image: .image-built
 # Push api-generator image remotely.
 .PHONY: push
 push: image
-	docker tag $(IMAGE_TAG) $(REMOTE_IMAGE)
-	docker push $(REMOTE_IMAGE)
+	${CONTAINER_ENGINE} tag $(IMAGE_TAG) $(REMOTE_IMAGE)
+	${CONTAINER_ENGINE} push $(REMOTE_IMAGE)
 
 # .image-built is a local token file to help Make determine the latest successful build.
 .image-built: Dockerfile
-	docker build ../.. -t $(IMAGE_TAG) -f Dockerfile
+	${CONTAINER_ENGINE} build ../.. -t $(IMAGE_TAG) -f Dockerfile
 	touch .image-built

test/imagebuilder/Makefile

@@ -14,6 +14,8 @@
 
 IMG = gcr.io/ml-pipeline-test/image-builder
 
+CONTAINER_ENGINE ?= docker
+
 # List any changed  files. We only include files in the notebooks directory.
 # because that is the code in the docker image.
 # In particular we exclude changes to the ksonnet configs.
@@ -31,17 +33,17 @@ all: build
 # To build without the cache set the environment variable
 # export DOCKER_BUILD_OPTS=--no-cache
 build:
-	docker build ${DOCKER_BUILD_OPTS} -t $(IMG):$(TAG) . \
+	${CONTAINER_ENGINE} build ${DOCKER_BUILD_OPTS} -t $(IMG):$(TAG) . \
            --label=git-verions=$(GIT_VERSION)
-	docker tag $(IMG):$(TAG) $(IMG):latest
+	${CONTAINER_ENGINE} tag $(IMG):$(TAG) $(IMG):latest
 	echo Built $(IMG):$(TAG)
 	echo Built $(IMG):latest
 
 # Build but don't attach the latest tag. This allows manual testing/inspection of the image
 # first.
 push: build
 	gcloud auth configure-docker
-	docker push $(IMG):$(TAG)
+	${CONTAINER_ENGINE} push $(IMG):$(TAG)
 	echo Pushed $(IMG):$(TAG)
 
 push-latest: push

test/release/Makefile

@@ -14,6 +14,8 @@
 
 REMOTE=gcr.io/ml-pipeline-test/release@sha256:ed1a4dbe536e7e161ad0d846b5681aacc0e0e7f285985cb1808c5c8987bcfeb0
 
+CONTAINER_ENGINE ?= docker
+
 .PHONY: release
 release:
 	# Usage: TAG=<TAG> BRANCH=<BRANCH> make release
@@ -33,17 +35,17 @@ release-in-place:
 # Build a release image locally using docker.
 .PHONY: build
 build:
-	docker build -t $(REMOTE) - < Dockerfile.release
+	${CONTAINER_ENGINE} build -t $(REMOTE) - < Dockerfile.release
 
 # Push locally built release image to remote container registry,
 # so that others can use the new image next time.
 .PHONY: push
 push: build
 	# Only some maintainers have access to push,
 	# contact @chensun @zijianjoy if you have any needs.
-	docker push $(REMOTE)
+	${CONTAINER_ENGINE} push $(REMOTE)
 
 # Run the docker image interactively in shell as current user.
 .PHONY: dev
 dev:
-	docker run -it -u $$(id -u):$$(id -g) $(REMOTE) /bin/bash
+	${CONTAINER_ENGINE} run -it -u $$(id -u):$$(id -g) $(REMOTE) /bin/bash