|author||Solomon Hykes <email@example.com>||Sat Mar 16 12:35:43 2013 -0700|
|committer||Solomon Hykes <firstname.lastname@example.org>||Sat Mar 16 12:35:43 2013 -0700|
Merge pull request #100 from silas/readme Fix coffee typo
It encapsulates heterogeneous payloads in Standard Containers, and runs them on any server with strong guarantees of isolation and repeatability.
Is is a great building block for automating distributed systems: large-scale web deployments, database clusters, continuous deployment systems, private PaaS, service-oriented architectures, etc.
Heterogeneous payloads: any combination of binaries, libraries, configuration files, scripts, virtualenvs, jars, gems, tarballs, you name it. No more juggling between domain-specific tools. Docker can deploy and run them all.
Any server: docker can run on any x64 machine with a modern linux kernel - whether it's a laptop, a bare metal server or a VM. This makes it perfect for multi-cloud deployments.
Isolation: docker isolates processes from each other and from the underlying host, using lightweight containers.
Repeatability: because containers are isolated in their own filesystem, they behave the same regardless of where, when, and alongside what they run.
Filesystem isolation: each process container runs in a completely separate root filesystem.
Resource isolation: system resources like cpu and memory can be allocated differently to each process container, using cgroups.
Network isolation: each process container runs in its own network namespace, with a virtual interface and IP address of its own.
Copy-on-write: root filesystems are created using copy-on-write, which makes deployment extremeley fast, memory-cheap and disk-cheap.
Logging: the standard streams (stdout/stderr/stdin) of each process container is collected and logged for real-time or batch retrieval.
Change management: changes to a container's filesystem can be committed into a new image and re-used to create more containers. No templating or manual configuration required.
Interactive shell: docker can allocate a pseudo-tty and attach to the standard input of any container, for example to run a throwaway interactive shell.
Under the hood, Docker is built on the following components:
AUFS, a powerful union filesystem with copy-on-write capabilities;
The Go programming language;
lxc, a set of convenience scripts to simplify the creation of linux containers.
Currently, Docker can be installed with Vagrant both on your localhost with VirtualBox as well as on Amazon EC2. Vagrant 1.1 is required for EC2, but deploying is as simple as:
$ export AWS_ACCESS_KEY_ID=xxx \ AWS_SECRET_ACCESS_KEY=xxx \ AWS_KEYPAIR_NAME=xxx \ AWS_SSH_PRIVKEY=xxx $ vagrant plugin install vagrant-aws $ vagrant up --provider=aws
The environment variables are:
AWS_ACCESS_KEY_ID- The API key used to make requests to AWS
AWS_SECRET_ACCESS_KEY- The secret key to make AWS API requests
AWS_KEYPAIR_NAME- The name of the keypair used for this EC2 instance
AWS_SSH_PRIVKEY- The path to the private key for the named keypair
For VirtualBox, you can simply ignore setting any of the environment variables and omit the
provider flag. VirtualBox is still supported with Vagrant <= 1.1:
$ vagrant up
sudo apt-get install lxc wget bsdtar curl
wget http://get.docker.io/builds/$(uname -s)/$(uname -m)/docker-master.tgz tar -xf docker-master.tgz
cd docker-master sudo ./docker run -a -i -t base /bin/bash
Consider adding docker to your
PATH for simplicity.
Right now, the officially supported distributions are:
Docker probably works on other distributions featuring a recent kernel, the AUFS patch, and up-to-date lxc. However this has not been tested.
# Download a base image docker import base # Run an interactive shell in the base image, # allocate a tty, attach stdin and stdout docker run -a -i -t base /bin/bash
# Run docker in daemon mode (docker -d || echo "Docker daemon already running") & # Start a very useful long-running process JOB=$(docker run /bin/sh -c "while true; do echo Hello world!; sleep 1; done") # Collect the output of the job so far docker logs $JOB # Kill the job docker kill $JOB
# Expose port 4444 of this container, and tell netcat to listen on it JOB=$(docker run -p 4444 base /bin/nc -l -p 4444) # Which public port is NATed to my container? PORT=$(docker port $JOB 4444) # Connect to the public port via the host's public address echo hello world | nc $(hostname) $PORT # Verify that the network connection worked echo "Daemon received: $(docker logs $JOB)"
Docker defines a unit of software delivery called a Standard Container. The goal of a Standard Container is to encapsulate a software component and all its dependencies in a format that is self-describing and portable, so that any compliant runtime can run it without extra dependency, regardless of the underlying machine and the contents of the container.
The spec for Standard Containers is currently work in progress, but it is very straightforward. It mostly defines 1) an image format, 2) a set of standard operations, and 3) an execution environment.
A great analogy for this is the shipping container. Just like Standard Containers are a fundamental unit of software delivery, shipping containers (http://bricks.argz.com/ins/7823-1/12) are a fundamental unit of physical delivery.
Just like shipping containers, Standard Containers define a set of STANDARD OPERATIONS. Shipping containers can be lifted, stacked, locked, loaded, unloaded and labelled. Similarly, standard containers can be started, stopped, copied, snapshotted, downloaded, uploaded and tagged.
Just like shipping containers, Standard Containers are CONTENT-AGNOSTIC: all standard operations have the same effect regardless of the contents. A shipping container will be stacked in exactly the same way whether it contains Vietnamese powder coffee or spare Maserati parts. Similarly, Standard Containers are started or uploaded in the same way whether they contain a postgres database, a php application with its dependencies and application server, or Java build artifacts.
Both types of containers are INFRASTRUCTURE-AGNOSTIC: they can be transported to thousands of facilities around the world, and manipulated by a wide variety of equipment. A shipping container can be packed in a factory in Ukraine, transported by truck to the nearest routing center, stacked onto a train, loaded into a German boat by an Australian-built crane, stored in a warehouse at a US facility, etc. Similarly, a standard container can be bundled on my laptop, uploaded to S3, downloaded, run and snapshotted by a build server at Equinix in Virginia, uploaded to 10 staging servers in a home-made Openstack cluster, then sent to 30 production instances across 3 EC2 regions.
Because they offer the same standard operations regardless of content and infrastructure, Standard Containers, just like their physical counterpart, are extremely well-suited for automation. In fact, you could say automation is their secret weapon.
Many things that once required time-consuming and error-prone human effort can now be programmed. Before shipping containers, a bag of powder coffee was hauled, dragged, dropped, rolled and stacked by 10 different people in 10 different locations by the time it reached its destination. 1 out of 50 disappeared. 1 out of 20 was damaged. The process was slow, inefficient and cost a fortune - and was entirely different depending on the facility and the type of goods.
Similarly, before Standard Containers, by the time a software component ran in production, it had been individually built, configured, bundled, documented, patched, vendored, templated, tweaked and instrumented by 10 different people on 10 different computers. Builds failed, libraries conflicted, mirrors crashed, post-it notes were lost, logs were misplaced, cluster updates were half-broken. The process was slow, inefficient and cost a fortune - and was entirely different depending on the language and infrastructure provider.
There are 17 million shipping containers in existence, packed with every physical good imaginable. Every single one of them can be loaded on the same boats, by the same cranes, in the same facilities, and sent anywhere in the World with incredible efficiency. It is embarrassing to think that a 30 ton shipment of coffee can safely travel half-way across the World in less time than it takes a software team to deliver its code from one datacenter to another sitting 10 miles away.
With Standard Containers we can put an end to that embarrassment, by making INDUSTRIAL-GRADE DELIVERY of software a reality.