Chris Hines

Video: What’s New in Docker Swarm 1.1

Last week we released our Docker Datacenter solution to the world, bringing container management and deployment to the enterprise with an on-premises Containers-as-a-Service platform. Universal Control Plane, part of the Docker Datacenter solution, enables IT operations teams to manage and deploy their dockerized distributed application anywhere, from behind the firewall.

But did you know that UCP sits on top of Docker Swarm? It uses Swarm to create and manage clusters, and because of this, can utilize the very same Docker APIs. Swarm is our production-ready orchestration solutions that serves as a clustering tool for Docker engines.
docker_swarm
A few weeks back we released our latest version of Swarm, version 1.1. This release included awesome new features like:

  1. Experimental support for rescheduling containers when a node fails
  2. Support for running images from private repositories
  3. Guides for getting started with Swarm with real-world use cases
  4. Better node management: Allowing users to see errors from nodes that fail to join clusters

To showcase these new features, we hosted a 1 hour webinar to highlight what the new updates mean for enterprise IT operations teams. Mike Coleman, Technical Marketing Engineer at Docker, was the main presenter. Watch the webinar below.


 

 

Here are some of the questions we received during the live presentation:

Q: Can you explain how to run Swarm master with HA?

All of this is very well documents within the Swarm docs.

Q: If we send 10 “docker-runs” can the Swarm manager run some of them? Or are they run by the other nodes?

They are run by the other nodes. The Swarm Manager just redirects traffic. Technically, there’s nothing prohibiting you from joining your master, but not best practice.

Q. Is there a Swarm container running on each node?

Yes, they are not shown in a normal Docker-ps, they are hidden. But yes there is one running on each node.

Q. Is there a recommended operating system?

No, we can run it on all of our supported OSs. As long as it’s running a Docker engine we use Swarm. Its part of the Docker platform. So RHEL, CentOS and Ubuntu. But really there isn’t a preferable one.

Q: Regarding this demo, can we follow the same instructions for MS Azure instead of AWS?

Yes you can, as long as you follow the same instructions you will be fine. Swarm is built for flexibility and we see that some enterprises have adopted multi-cloud environments (i.e AWS + Azure). Our goal is to enable this and allow cloud portability for Docker containers.

Q: Is it essential to have Docker network created in order for Swarm to work?

No it is not, the network just makes it easier. If you didn’t have an overlay network you would have to use links to make containers talk to each other, which is way more cumbersome.
Swarmnado-357x627-30 (1)

 

Be sure to save the date for SwarmWeek, which begins on March 7th!

 
Here are some resources for additional information:

 


 

Learn More about Docker

, , , , , , , , ,

Chris Hines

Video: What’s New in Docker Swarm 1.1


Leave a Reply

Get the Latest Docker News by Email

Docker Weekly is a newsletter with the latest content on Docker and the agenda for the upcoming weeks.