Mohammad Umair Khan
Trainer, Mentor, Foo, Bar !
Cherry's tale
from manual to automation
Cherry is a standard
"Let's build an MVP"
- Have changing teams
- Require fast development and deployment cycles
- Want something that doesn't break
- Grow really fast once they hit it off
Often ...
Highlights
May, 2017
Mar, 2018
version: "3"
services:
app:
container_name: cherry-api
build:
context: .
dockerfile: Dockerfile
...
orient:
image: orientdb:2.2.28-spatial
container_name: orientdb
volumes:
- "orientdb:/orientdb/databases"
- "orientdb_backup:/orientdb/backup"
ports:
- 2424:2424
- 2480:2480
env_file:
- ./env/development
volumes:
orientdb:
orientdb_backup:
Standard docker-compose.yml
Host OS
Docker Engine
docker network (foldername_default)
Node App
OrientDB
$ docker-compose up
run
- Environment woes
- Versioning woes
- Setup time slack
Create image
Push image
SSH to remote
Pull image
Run compose
Create image
Push image
SSH to remote
Pull image
Run compose
#!/bin/sh
Create new task definition
what to deploy
Create Service
Deployment
Deployment
Where to deploy
(1) Build
(2) Push
(3) Create new task definition
(5) Update service
(6) Deploy
(6) Deploy
app.cherry.expert
api.cherry.expert
- Zero down time deployments
- No SSH deployments
- Auto recovery
- Scale horizontally
We follow a hybrid agile methodology
Requirements and solutions evolve through the collaborative effort of self-organizing cross-functional teams and their customer/end users
Which also means
- Requirements being added on the fly
- Sprints having more bandwidth then all SEA-ME-WEs combined
- Having 4 product owners because the real one is on vacation
- Whatever it takes, velocity should improve
Node App
OrientDB
JMeter
Cherry
#!/bin/sh
1) Create PR
2) Invokes hook
3) Run tests
4a) Failed/Rejected
4b) Passed/Merged
10
Cherry's tale: From manual to automation
By Mohammad Umair Khan
