Skip to content

Latest commit

 

History

History
205 lines (143 loc) · 9.13 KB

faqs.md

File metadata and controls

205 lines (143 loc) · 9.13 KB
Raw
description keywords title toc_max
Frequently asked questions
aws faqs
Docker for AWS frequently asked questions (FAQ)
2

Stable and edge channels

Two different download channels are available for Docker for AWS:

  • The stable channel provides a general availability release-ready deployment for a fully baked and tested, more reliable cluster. The stable version of Docker for AWS comes with the latest released version of Docker Engine. The release schedule is synched with Docker Engine releases and hotfixes. On the stable channel, you can select whether to send usage statistics and other data.

  • The edge channel provides a deployment with new features we are working on, but is not necessarily fully tested. It comes with the experimental version of Docker Engine. Bugs, crashes, and issues are more likely to occur with the edge cluster, but you get a chance to preview new functionality, experiment, and provide feedback as the deployment evolve. Releases are typically more frequent than for stable, often one or more per month. Usage statistics and crash reports are sent by default. You do not have the option to disable this on the edge schannel.

Can I use my own AMI?

No, at this time we only support the default Docker for AWS AMI.

How can I use Docker for AWS with an AWS account in an EC2-Classic region?

If you have an AWS account that was created before December 4th, 2013 you have what is known as an EC2-Classic account on regions where you have previously deployed resources. EC2-Classic accounts don't have default VPC's or the associated subnets, etc. This causes a problem when using our CloudFormation template because we are using the Fn:GetAZs function they provide to determine which availability zones you have access to. When used in a region where you have EC2-Classic, this function returns all availability zones for a region, even ones you don't have access to. When you have an EC2-VPC account, it returns only the availability zones you have access to.

This causes an error like the following:

"Value (us-east-1a) for parameter availabilityZone is invalid. Subnets can currently only be created in the following availability zones: us-east-1d, us-east-1c, us-east-1b, us-east-1e."

If you have an EC2-Classic account, and you don't have access to the a and b availability zones for that region.

There isn't anything we can do right now to fix this issue. We have contacted Amazon to provide a solution.

How to tell if you are in the EC2-Classic region.

This AWS documentation page describes how you can tell if you have EC2-Classic, EC2-VPC or both.

Possible fixes to the EC2-Classic region issue:

There are a few workarounds that you can try to get Docker for AWS up and running for you.

  1. Create your own VPC, then install Docker for AWS with a pre-existing VPC.
  2. Use a region that doesn't have EC2-Classic. The most common region with this issue is us-east-1. So try another region, us-west-1, us-west-2, or the new us-east-2. These regions should be set up with EC2-VPC and the issue shouldn't occur.
  3. Create an new AWS account, all new accounts are setup using EC2-VPC and do not have this problem.
  4. Contact AWS support to convert your EC2-Classic account to a EC2-VPC account. For more information checkout the following answer for "Q. I really want a default VPC for my existing EC2 account. Is that possible?" on https://aws.amazon.com/vpc/faqs/#Default_VPCs

Helpful links:

Can I use my existing VPC?

Yes, see install Docker for AWS with a pre-existing VPC for more info.

Recommended VPC and subnet setup

VPC

  • CIDR: 172.31.0.0/16
  • DNS hostnames: yes
  • DNS resolution: yes
  • DHCP option set: DHCP Options (Below)

Internet gateway

  • VPC: VPC (above)

DHCP option set

  • domain-name: ec2.internal
  • domain-name-servers: AmazonProvidedDNS

Subnet1

  • CIDR: 172.31.16.0/20
  • Auto-assign public IP: yes
  • Availability-Zone: A

Subnet2

  • CIDR: 172.31.32.0/20
  • Auto-assign public IP: yes
  • Availability-Zone: B

Subnet3

  • CIDR: 172.31.0.0/20
  • Auto-assign public IP: yes
  • Availability-Zone: C

Route table

  • Destination CIDR block: 0.0.0.0/0
  • Subnets: Subnet1, Subnet2, Subnet3
Subnet note:

If you are using the 10.0.0.0/16 CIDR in your VPC. When you create a docker network, you need to pick a subnet (using docker network create —subnet option) that doesn't conflict with the 10.0.0.0 network.

Which AWS regions does this work with?

Docker for AWS should work with all regions except for AWS US Gov Cloud (us-gov-west-1) and AWS China, which are a little different than the other regions.

How many Availability Zones does Docker for AWS use?

Docker for AWS determines the correct amount of Availability Zone's to use based on the region. In regions that support it, we use 3 Availability Zones, and 2 for the rest of the regions. We recommend running production workloads only in regions that have at least 3 Availability Zones.

What do I do if I get KeyPair error on AWS?

As part of the prerequisites, you need to have an SSH key uploaded to the AWS region you are trying to deploy to. For more information about adding an SSH key pair to your account, refer to the Amazon EC2 Key Pairs docs.

Where are my container logs?

All container logs are aggregated within AWS CloudWatch.

Best practice to deploy a large cluster

When deploying a cluster of more than 20 workers, it can take a very long time for AWS to deploy all of the instances (1+hrs). It is best to deploy a cluster of 20 workers then scale it up in the Auto-Scaling Group (ASG) once it's been deployed.

Benchmark of 3 Managers (m4.large) + 200 workers (t2.medium):

  • Deploying (~3.1hrs)
    • Deployment: 3 Managers + 200 workers = ~190mins
  • Scaling (~35mins)
    • Deployment: 3 Managers + 20 workers = ~20mins
    • Scaling: 20 workers -> 200 workers via ASG = ~15mins

Note: During a Stack upgrade, you need to match the Auto-Scaling Group worker count, otherwise AWS scales it back down (aka type 200 workers in the input box)

Where do I report problems or bugs?

Send an email to [email protected] or post to the Docker for AWS GitHub repositories.

In AWS, if your stack is misbehaving, run the following diagnostic tool from one of the managers - this collects your docker logs and send them to Docker:

$ docker-diagnose
OK hostname=manager1
OK hostname=worker1
OK hostname=worker2
Done requesting diagnostics.
Your diagnostics session ID is 1234567890-xxxxxxxxxxxxxx
Please provide this session ID to the maintainer debugging your issue.

Note: Your output may be slightly different from the above, depending on your swarm configuration.

Metrics

Docker for AWS sends anonymized minimal metrics to Docker (heartbeat). These metrics are used to monitor adoption and are critical to improve Docker for AWS.

How do I run administrative commands?

By default when you SSH into a manager, you are logged in as the regular username: docker - It is possible however to run commands with elevated privileges by using sudo. For example to ping one of the nodes, after finding its IP via the Azure/AWS portal, such as 10.0.0.4, you could run:

$ sudo ping 10.0.0.4

Note: Access to Docker for AWS and Azure happens through a shell container that itself runs on Docker.

What are the Editions containers running after deployment?

In order for our editions to deploy properly and for load balancer integrations to happen, we run a few containers. They are as follow:

Container name Description
init Sets up the swarm and makes sure that the stack came up properly. (checks manager+worker count).
shell This is our shell/ssh container. When you SSH into an instance, you're actually in this container.
meta Assist in creating the swarm cluster, giving privileged instances the ability to join the swarm.
l4controller Listens for ports exposed at the docker CLI level and opens them in the load balancer.

How do I uninstall Docker for AWS?

You can remove the Docker for AWS setup and stacks through the AWS Console{: target="blank" class=""} on the CloudFormation page. See Uninstalling or removing a stack.