CODEWITHSUNDEEP
amazon-web-servicesamazon-ec2dockeramazon-ecs
Alex L
Alex L

Reputation: 651

How do I run an nginx reverse proxy container on ECS using arguments

I have a slightly modified version of the container here:

http://jasonwilder.com/blog/2014/03/25/automated-nginx-reverse-proxy-for-docker/

To run it locally I use:

$ docker run -d -p 80:80 -v /var/run/docker.sock:/tmp/docker.sock -t jwilder/nginx-proxy

And run my other containers similar to below:

$ docker run -e VIRTUAL_HOST=my-domain.local -t -d my-repo/site-name

I'm having trouble converting this to a Task Definition on ECS.

Task definition

I have two containers:

nginx-reverse-proxy

static-site

I have a cluster with one ECS instance and a server with one task, defined above. This keeps cycling and fails with "STOPPED (Essential container in task exited)".

Now I assume I need to setup volumes but all the examples I can find have a path and a name and I can't see how I can convert this "/var/run/docker.sock:/tmp/docker.sock" into any fields available.

Upvotes: 3

Views: 4845

Answers (1)

briancurt
briancurt

Reputation: 132

I'm quite late, but I'm going to answer anyway for people who come across this post in the future.

You have answered yourself. It fails because you are not sharing the Docker socket from the host, which is essential to the nginx proxy image.

Go to your Task Definition and add a new Volume. Name it socket and specify the host path /var/run/docker.sock. Then edit your nginx proxy container and under Mount Points choose socket as a source volume and /tmp/docker.sock as the container path. Mark it as read only for security reasons too. Leave the Command section for this container with the default value.

Deploy your new revision and it should work now.

Upvotes: 4

Related Questions