Adding docker config

.gitignore

@@ -169,3 +169,4 @@ data/**/*.json
 .vscode
 *.csv
 test-results/*
+manifest*.yml

Dockerfile

@@ -0,0 +1,34 @@
+FROM node:dubnium-slim
+
+ARG NODE_ENV
+ENV NODE_ENV=${NODE_ENV}
+
+ARG ELASTICSEARCH_SERVICE_NAME
+ENV ELASTICSEARCH_SERVICE_NAME=${ELASTICSEARCH_SERVICE_NAME}
+
+ARG LOGGER_LEVEL
+ENV LOGGER_LEVEL=${LOGGER_LEVEL}
+
+ARG GET_REMOTE_METADATA
+ENV GET_REMOTE_METADATA=${GET_REMOTE_METADATA}
+
+ARG GET_GITHUB_DATA
+ENV GET_GITHUB_DATA=${GET_GITHUB_DATA}
+
+ARG REMOTE_METADATA_LOCATION
+ENV REMOTE_METADATA_LOCATION=${REMOTE_METADATA_LOCATION}
+
+ARG GITHUB_API_KEY
+ENV GITHUB_API_KEY=${GITHUB_API_KEY}
+
+ARG ES_URI
+ENV ES_URI=${ES_URI}
+
+WORKDIR /usr/src/app
+VOLUME "/usr/src/app"
+
+COPY package*.json ./
+
+RUN npm install
+
+CMD [ "npm", "start" ]

README.md

@@ -37,7 +37,10 @@ Before running any of the commands included in the package.json file there are s
 
 #### Docker
 
-To be added soon.
+A docker-compose file has been added for convience. To run a dockerized version of the harvester, the standard `npm install` must be ran first; the files are copied into the container from the host.
+Then, running `docker-compose up` will start the container.
+
+This `docker-compose` file includes an elasticsearch configuration running on port 9200.
 
 ### Configuring
 

docker-compose.yml

@@ -0,0 +1,35 @@
+version: '2.4'
+services:
+  harvester:
+    build: .
+    image: 'codegov/code-gov-harvester:latest'
+    container_name: 'code-gov-harvester'
+    env_file:
+      - ./.env
+    depends_on:
+      - elasticsearch
+    volumes:
+      - .:/usr/src/app
+    command: ["npm", "start"]
+    networks:
+    - esnet
+
+  elasticsearch:
+    image: "docker.elastic.co/elasticsearch/elasticsearch:5.6.9"
+    environment:
+      - xpack.security.enabled=false
+    ports:
+      - "9200:9200"
+      - "9300:9300"
+    container_name: 'code-gov-data'
+    healthcheck:
+      test: ["CMD-SHELL", "curl --silent --fail localhost:9200/_cluster/health || exit 1"]
+      interval: 30s
+      timeout: 30s
+      retries: 3
+    networks:
+    - esnet
+
+networks:
+  esnet:
+    driver: bridge

manifest.yml → manifest.yml.example

@@ -1,5 +1,5 @@
 applications:
-- name: code-gov-harvester
+- name: code-gov-harvester-v1.0.0
   memory: 200M
   instances: 1
   disk_quota: 200M
@@ -11,8 +11,8 @@ applications:
     CF_STAGING_TIMEOUT: 15
     CF_STARTUP_TIMEOUT: 15
     LOGGER_LEVEL: info
-    GITHUB_TOKEN: not-a-token
     GITHUB_AUTH_TYPE: token
+    GITHUB_TOKEN: NOT_A_TOKEN
     GET_REMOTE_METADATA: true
     REMOTE_METADATA_LOCATION: https://raw.githubusercontent.com/GSA/code-gov-data/master/agency_metadata.json
     GET_GITHUB_DATA: true
@@ -35,4 +35,4 @@ applications:
   services:
   - code_gov_elasticsearch
   stack: cflinuxfs3
-  command: npm start
+  command: npm start

package.json

@@ -1,6 +1,6 @@
 {
   "name": "code-gov-harvester",
-  "version": "0.0.3",
+  "version": "1.0.3",
   "description": "Code.gov code.json harvester and data manipulation tool",
   "scripts": {
     "debug": "node --inspect ./scripts/index/index.js",