Data Focused Docker Clustering. Docker Hamburg

Post on 02-Jul-2015

237 views 0 download

description

Handling persistent data with Docker using Flocker

transcript

Docker HamburgThe upcoming talk will be presented in English

Data focused DockerclusteringMarcus Hughes

What was HybridCluster?

• Full stack LAMP platform

• FreeBSD Chroot containers

• ZFS

• Auto scaling

• Failover

• “AwesomeProxy”

Docker

• Containers!

• Application level containerisation

• Data in “volumes”

• Lack of data portability

Docker containers are portable

Introduction to Flocker

• Open source volume manager

• Multi-node “cluster”

• 0.1 Released on 13th August (0.2 due on 2nd)

Flocker

Proxy layer

• Moving an application can be disruptive

• Application port responds on any node

• Routes reconfigured on migration

ZFS

• Similarities to Docker containers

• ZFS on Linux

• Resilient

• Continuous replication

Migration

• Container stops

• Volumes (ZFS dataset) is sent

• Routes altered

• Container re-started on new host

0.2

• Incremental send

Application configuration

"version": 1

"applications":

"mongodb-volume-example":

"image": "clusterhq/mongodb"

"ports":

- "internal": 27017

"external": 27017

"volume":

"mountpoint": "/data/db"

Deployment configuration

"version": 1

"nodes":

"172.16.255.250": ["mongodb-volume-example"]

"172.16.255.251": []

Deployment configuration

flocker-deploy deployment.yml application.yml

DemoLets see what these configuration files do!

Real world uses

• Moving applications to production

• Replacing servers

• Scalability

Future

• Clone datasets

The FutureIdeas from our past and possible directions

Contribute!

• GitHub https://github.com/ClusterHQ/flocker

• Email Marcus@ClusterHQ.com

Questions?