From 17a9920f2f69b034212d34c0871f7967af87af5a Mon Sep 17 00:00:00 2001
From: Eric Brown <ericbrown@khanacademy.org>
Date: Mon, 3 Oct 2022 13:51:23 -0700
Subject: [PATCH] INFRA-8460 Initial kabuild container (#65)

Note: Naming and location of this container is not final.
It may be better located in webapp. Nevertheless, it is
useful now in several ways.

* Try to do everything that linux-setup.sh does
* Enough debug tools that this container is useful for debugging things

Issue: https://khanacademy.atlassian.net/browse/INFRA-8460

Test plan:

Run the following commands:
- make build
- make run_dirty

Check that ~/khan contains local ~/khan and that git works.
This is likely very slow on a mac, so be careful.

Ideally, try a clean checkout of webapp somewhere followed by
"make deps". This is probably best tested on a linux machine.
---
 .gitignore                                 |   1 +
 containers/kabuild/Dockerfile              | 153 +++++++++++++++++++++
 containers/kabuild/Dockerfile.dockerignore |   4 +
 containers/kabuild/Makefile                |  71 ++++++++++
 containers/kabuild/README.md               |  58 ++++++++
 containers/kabuild/k8s-pvc.yaml            |  26 ++++
 containers/kabuild/k8s.yaml                |  23 ++++
 containers/kabuild/skaffold.yaml           |  26 ++++
 skaffold.yaml                              |   6 +
 9 files changed, 368 insertions(+)
 create mode 100644 containers/kabuild/Dockerfile
 create mode 100644 containers/kabuild/Dockerfile.dockerignore
 create mode 100644 containers/kabuild/Makefile
 create mode 100644 containers/kabuild/README.md
 create mode 100644 containers/kabuild/k8s-pvc.yaml
 create mode 100644 containers/kabuild/k8s.yaml
 create mode 100644 containers/kabuild/skaffold.yaml
 create mode 100644 skaffold.yaml

diff --git a/.gitignore b/.gitignore
index 8233462..253f82c 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,2 +1,3 @@
 /**/packer_cache/
 /**/builds/
+.vscode
diff --git a/containers/kabuild/Dockerfile b/containers/kabuild/Dockerfile
new file mode 100644
index 0000000..edb3c0c
--- /dev/null
+++ b/containers/kabuild/Dockerfile
@@ -0,0 +1,153 @@
+# This is meant to be a base image for building (and possibly running) webapp.
+# It implements most of linux-setup.sh.
+
+# Warning: We're using a number of "deprecated" packages and moving past 22.04
+# is likely non-trivial.
+FROM ubuntu:22.04
+
+# TODO: Use linux-setup.sh for everything.
+# Warning: I hit a number of snags trying to use linux-setup.sh directly.
+
+RUN apt-get update && apt-get install -y \
+    curl \
+    git \
+    jq \
+    make \
+    curl \
+    unzip \
+    net-tools \
+    telnet \
+    wget \
+    screen \
+    tree \
+    netcat \
+    strace \
+    tcpdump \
+    tcpflow \
+    sudo \
+    software-properties-common \
+    apt-transport-https \
+    ca-certificates \
+    gnupg
+
+# We're in a container. We aren't root just to avoid stupid mistakes.
+# But we want to be root easily if needed.
+RUN echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers
+
+# TODO: Lock down google-sdk version as we do elsewhere.
+
+# Install google-sdk repo
+RUN curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add -
+RUN echo "deb https://packages.cloud.google.com/apt cloud-sdk main" >> /etc/apt/sources.list.d/google-cloud-sdk.list
+# Install golang repo
+RUN add-apt-repository -y ppa:longsleep/golang-backports
+RUN add-apt-repository -y -r ppa:chris-lea/node.js
+RUN add-apt-repository -y ppa:git-core/ppa
+
+RUN wget -O- https://deb.nodesource.com/gpgkey/nodesource.gpg.key | sudo apt-key add -
+RUN echo deb https://deb.nodesource.com/node_16.x `lsb_release -c -s` main >> /etc/apt/sources.list.d/nodesource.list
+RUN echo deb-src https://deb.nodesource.com/node_16.x `lsb_release -c -s` main >> /etc/apt/sources.list.d/nodesource.list
+RUN chmod a+rX /etc/apt/sources.list.d/nodesource.list
+RUN echo "Package: nodejs" >> /etc/apt/preferences.d/nodejs
+RUN echo "Pin: version 16.*" >> /etc/apt/preferences.d/nodejs
+RUN echo "Pin-Priority: 999" >> /etc/apt/preferences.d/nodejs
+
+# Update indexes
+RUN apt-get update
+
+# Install google cloud sdk
+RUN apt-get install -y \
+    google-cloud-sdk \
+    openjdk-11-jdk \
+    nodejs \
+    watchman
+
+# TODO: Get version from common location
+# WARNING: We REALLY WANT to avoid requring make and/or break standalone skaffold.
+ENV DESIRED_GO_VERSION=1.16
+RUN apt-get install -y golang-${DESIRED_GO_VERSION}
+RUN cp -sf /usr/lib/go-${DESIRED_GO_VERSION}/bin/* /usr/local/bin/
+
+# Python2
+ENV VENV=.virtualenv/khan27
+RUN apt-get install -y python2-dev python-setuptools
+RUN ln -s /usr/bin/python2 /usr/bin/python
+WORKDIR /tmp
+RUN curl https://bootstrap.pypa.io/pip/2.7/get-pip.py --output get-pip.py
+RUN python2 get-pip.py pip==19.3.1
+RUN rm -f get-pip.py
+WORKDIR /root
+RUN pip install virtualenv==20.0.23
+RUN pip install http://sourceforge.net/projects/pychecker/files/pychecker/0.8.19/pychecker-0.8.19.tar.gz/download
+RUN virtualenv -q --python="$(which python2)" --always-copy "/root/${VENV}"
+RUN /root/${VENV}/bin/pip install -U "pip<20" setuptools
+
+RUN apt-get install -y \
+        libfreetype6 libfreetype6-dev libpng-dev libjpeg-dev \
+        imagemagick \
+        libxslt1-dev \
+        libyaml-dev \
+        libncurses-dev libreadline-dev \
+        nginx \
+        redis-server \
+        libnss3-tools \
+        python3-pip
+
+RUN pip3 install -q pipenv
+RUN npm install -g yarn
+
+# protoc
+RUN mkdir -p /tmp/protoc
+RUN wget -O /tmp/protoc/protoc-3.4.0.zip https://github.com/google/protobuf/releases/download/v3.4.0/protoc-3.4.0-linux-x86_64.zip
+WORKDIR /tmp/protoc
+RUN unzip -q protoc-3.4.0.zip
+RUN install -m755 ./bin/protoc /usr/local/bin
+RUN rm -rf /usr/local/include/google/protobuf
+RUN mkdir -p /usr/local/include/google
+RUN mv ./include/google/protobuf /usr/local/include/google/
+RUN chmod -R a+rX /usr/local/include/google/protobuf
+WORKDIR /root
+RUN rm -rf /tmp/protoc
+
+WORKDIR /root
+
+ENV GROUP_NAME=khandev
+ENV USER_NAME=khandev
+RUN addgroup --gid 1001 ${GROUP_NAME}
+# To set password to "kabuild", add -p padco2M.JnJHY
+RUN useradd -ms /bin/bash --uid 1001 --gid 1001 -G sudo ${USER_NAME}
+
+RUN mkdir /tmp/fastly
+RUN cd /tmp/fastly && \
+    curl -LO https://github.com/fastly/cli/releases/download/v3.3.0/fastly_3.3.0_linux_amd64.deb && \
+    apt install ./fastly_3.3.0_linux_amd64.deb
+RUN rm -rf /tmp/fastly
+
+USER ${USER_NAME}:${GROUP_NAME}
+WORKDIR /home/${USER_NAME}
+
+RUN mkdir -m 700 .ssh
+RUN mkdir khan
+RUN ssh-keyscan github.com >> ~/.ssh/known_hosts
+RUN virtualenv -q --python="$(which python2)" --always-copy "/home/${USER_NAME}/${VENV}"
+RUN /home/${USER_NAME}/${VENV}/bin/pip install -U "pip<20" setuptools
+COPY --chown=${USER_NAME}:${GROUP_NAME} .profile.khan .profile.khan
+COPY --chown=${USER_NAME}:${GROUP_NAME} .bash_profile.khan .bash_profile.khan
+COPY --chown=${USER_NAME}:${GROUP_NAME} .bashrc.khan .bashrc.khan
+COPY --chown=${USER_NAME}:${GROUP_NAME} bash_profile.default .bash_profile
+COPY --chown=${USER_NAME}:${GROUP_NAME} profile.default .profile
+COPY --chown=${USER_NAME}:${GROUP_NAME} git-completion.bash git-completion.bash
+COPY --chown=${USER_NAME}:${GROUP_NAME} .gitconfig.khan .gitconfig
+COPY --chown=${USER_NAME}:${GROUP_NAME} . /home/${USER_NAME}/khan/devtools/khan-dotfiles/
+
+COPY --chown=${USER_NAME}:${GROUP_NAME} bashrc.default bashrc.default
+RUN cat bashrc.default .bashrc > .bashrc.new
+RUN mv .bashrc.new .bashrc
+RUN rm -f bashrc.default
+
+RUN mkdir -p /tmp/rust
+RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs --output /tmp/rust/rustup-init.sh
+RUN bash /tmp/rust/rustup-init.sh -y --profile default --no-modify-path
+RUN rm -rf /tmp/rust
+
+CMD tail -f /dev/null
diff --git a/containers/kabuild/Dockerfile.dockerignore b/containers/kabuild/Dockerfile.dockerignore
new file mode 100644
index 0000000..acb5229
--- /dev/null
+++ b/containers/kabuild/Dockerfile.dockerignore
@@ -0,0 +1,4 @@
+.git
+containers/kabuild/k8s*.yaml
+containers/kabuild/Makefile
+*.md
diff --git a/containers/kabuild/Makefile b/containers/kabuild/Makefile
new file mode 100644
index 0000000..fcd0939
--- /dev/null
+++ b/containers/kabuild/Makefile
@@ -0,0 +1,71 @@
+REPO=gcr.io/khan-internal-services/services
+IMAGE=${REPO}/kabuild-image
+CURRENT_IMAGE_TAG:=$(shell skaffold build -q --dry-run | jq '.builds[0].tag' | tr -d '"')
+ifeq (${VER},)
+CLEAN_IMAGE_TAG:=$(shell echo ${CURRENT_IMAGE_TAG} | sed 's/-dirty//')
+else
+CLEAN_IMAGE_TAG:=${IMAGE}:${VER}
+endif
+ifneq (${CURRENT_IMAGE_TAG},${CLEAN_IMAGE_TAG})
+PULL_POLICY=always
+else
+PULL_POLICY=missing
+endif
+
+KAROOT=~/khan
+
+.PHONY: help
+help:
+	@echo "  build: Build the container with cloudbuild"
+	@echo "  docker_build: Build the container locally with docker"
+	@echo "  repo_list: List the images we've pushed so far"
+	@echo "  run_dirty: Run the container most recently built (for debugging changes)"
+	@echo "  run: Run what we think is the most recent version of the container"
+	@echo "  run VER=...: Run a specific version of the container"
+	@echo "  current_tag: The current tag that is/was built"
+	@echo "  current image: Full image that is/was built"
+
+# The current image name (and tag)
+.PHONY: current_image
+current_image:
+	@echo ${CURRENT_IMAGE_TAG}
+
+# The current tag that will be built
+.PHONY: current_tag
+current_tag:
+	@echo ${CURRENT_IMAGE_TAG} | sed 's/.*://'
+
+# Run what we think is the most recent non-dirty image (most recent tag)
+.PHONY: run
+run:
+	@echo "Running ${CLEAN_IMAGE_TAG}"
+	@echo "Mapping private key ~/.ssh/id_rsa"
+	@echo "Mapping ~/khan"
+	@docker run --rm -it \
+		-v ~/.ssh/id_rsa:/home/kabuild/.ssh/id_rsa \
+		-v ${KAROOT}:/home/kabuild/khan \
+		${CLEAN_IMAGE_TAG} bash -l
+
+# Run the current image whether dirty or clean
+# This is the command to use when testing changes
+.PHONY: run_dirty
+run_dirty:
+	@echo "Running ${CURRENT_IMAGE_TAG}"
+	@echo "Mapping private key ~/.ssh/id_rsa"
+	@echo "Mapping ~/khan"
+	@docker run -it --rm --pull ${PULL_POLICY} \
+		-v ~/.ssh/id_rsa:/home/kabuild/.ssh/id_rsa \
+		-v ${KAROOT}:/home/kabuild/khan \
+		${CURRENT_IMAGE_TAG} bash -l
+
+.PHONY: docker_build
+docker_build:
+	skaffold build
+
+.PHONY: build
+build:
+	skaffold build -p cloudbuild
+
+.PHONY: repo_list
+repo_list:
+	gcloud container images list-tags ${IMAGE}
diff --git a/containers/kabuild/README.md b/containers/kabuild/README.md
new file mode 100644
index 0000000..1eb457c
--- /dev/null
+++ b/containers/kabuild/README.md
@@ -0,0 +1,58 @@
+# kabuild-image
+
+## Overview
+
+This is an experimental image meant to replace the "base" image in webapp. 
+It is purely ubuntu 22.04 with the goal to encapsulate all developer tooling
+required to run webapp's "make deps". As tooling required to do that is
+defined mostly in khan-dotfiles, this is where we are defining the image.
+
+## Versioning
+
+We are using semantic versioning starting with 1.0.0. Versions are created
+by a developer manually creating a git-tag containing the semantic version.
+
+## Why skaffold
+
+skaffold is a framework for building and deployment that is a bit opionated.
+When using defaults (which aren't necessarily exactly the way Khan typically
+does things), it also simplifies aspects of build and deployment. Thus, we
+are attempting to use skaffold instead of calling docker (or even make in
+most cases).
+
+The only skaffold command we're really using here is "skaffold build"
+
+## Why NOT linux-setup.sh
+
+linux-setup.sh & linux-functions.sh could possibly be used instead of
+doing things explicity in the Dockerfile. But that was tried and there was
+quite a bit to debug. And in the long term, DevOps anticipates that all
+services are built with relatively simple containers instead of a mega
+kabuild kitchen-sink container like this one. Thus, the effort to debug
+scripts is minimal value. Others are welcome to put in the effort.
+
+## Building
+
+To build locally, run: skaffold build
+
+To build remotely, run: skaffold build -p cloudbuild
+
+## Tagging and building releases
+
+Once pushed to master, a developer should tag that version of master
+with an appropriate semantic version (i.e. 1.3.5), push the tag and 
+then run "make build" locally. That will build the semantic version
+and push it to the container repository.
+
+## Other Usage
+
+Please see "make help"
+
+## Remote Use
+
+TBD - i.e. Using "skaffold run" with an appropriate persistent
+volume, claim and appropriate ssh-key mapping outght to make remote
+development in kubernetes quite realistic. However, ideally this would
+be bootstrapped with another container that has the most recent webapp
+repo and run "make deps", etc. (This may be somewhat trivial with a
+webhook or nightly task.)
diff --git a/containers/kabuild/k8s-pvc.yaml b/containers/kabuild/k8s-pvc.yaml
new file mode 100644
index 0000000..8b5b334
--- /dev/null
+++ b/containers/kabuild/k8s-pvc.yaml
@@ -0,0 +1,26 @@
+kind: PersistentVolumeClaim
+apiVersion: v1
+metadata:
+  name: khandev-claim
+spec:
+  storageClassName: khandev-ssd-retain
+  resources:
+    requests:
+      storage: 100Gi
+  accessModes:
+    - ReadWriteOnce
+---
+# This storage class overrides standard-rwo to force Retain=true
+# Thus, data sticks around.
+apiVersion: storage.k8s.io/v1
+kind: StorageClass
+metadata:
+  name: khandev-ssd-retain
+provisioner: kubernetes.io/gce-pd
+volumeBindingMode: Immediate
+allowVolumeExpansion: true
+reclaimPolicy: Retain
+parameters:
+  type: pd-ssd
+  fstype: ext4
+  replication-type: none
diff --git a/containers/kabuild/k8s.yaml b/containers/kabuild/k8s.yaml
new file mode 100644
index 0000000..c71bf2f
--- /dev/null
+++ b/containers/kabuild/k8s.yaml
@@ -0,0 +1,23 @@
+apiVersion: v1
+kind: Pod
+metadata:
+  name: &service-name khandev
+spec:
+  volumes:
+  - name: khandev-volume
+    persistentVolumeClaim:
+      claimName: khandev-claim
+  containers:
+  - name: *service-name
+    image: kabuild-image
+    volumeMounts:
+    - mountPath: "/home/khandev/khan"
+      name: khandev-volume
+    resources:
+      requests:
+        memory: 512Mi
+        cpu: 250m
+      limits:
+        memory: 4096Mi
+        cpu: 1000m
+  restartPolicy: Never
diff --git a/containers/kabuild/skaffold.yaml b/containers/kabuild/skaffold.yaml
new file mode 100644
index 0000000..f10ba45
--- /dev/null
+++ b/containers/kabuild/skaffold.yaml
@@ -0,0 +1,26 @@
+apiVersion: skaffold/v2beta29
+kind: Config
+metadata:
+  name: kabuild
+build:
+  artifacts:
+  - image: kabuild-image
+    docker:
+      dockerfile: containers/kabuild/Dockerfile
+    context: ../..
+  tagPolicy:
+    gitCommit: {}
+deploy:
+  kubectl:
+    manifests:
+    - k8s.yaml
+portForward:
+- resourceType: Pod
+  port: 8088
+- resourceType: Pod
+  port: 2000
+profiles:
+- name: cloudbuild
+  build:
+    googleCloudBuild:
+      projectId: khan-internal-services
diff --git a/skaffold.yaml b/skaffold.yaml
new file mode 100644
index 0000000..01a18e9
--- /dev/null
+++ b/skaffold.yaml
@@ -0,0 +1,6 @@
+apiVersion: skaffold/v2beta29
+kind: Config
+metadata:
+  name: dotfiles
+requires:
+- path: containers/kabuild