$ conda create --name Oraclize-BioID
$ source activate Oraclize-BioID
$ pip install azure-cli==2.0.12
You will want to update azuredeploy.parameters.json before proceeding.
You can find replace ti-acs-swarm to your-app-acs-swarm. Please omit this change in any PRs.
$ az login
$ az acs list
$ az group create --name "ti-acs-swarm-rg" --location "southcentralus"
$ az group deployment create -g ti-acs-swarm-rg -n ti-acs-swarm --template-uri https://raw.githubusercontent.com/Azure/azure-quickstart-templates/master/101-acs-swarm/azuredeploy.json --parameters ./azure/azuredeploy.parameters.json
Before we proceed, please read this:
These instructions focus on tunneling TCP traffic over SSH.
You can also start an interactive SSH session with one of the internal cluster management systems,
but we don't recommend this. Working directly on an internal system risks inadvertent configuration changes.
# establish ssh tunnel for docker to use.
$ ssh -fNL 2375:localhost:2375 -p 2200 or13@ti-acs-swarmmgmt.southcentralus.cloudapp.azure.com
# if you encounter errors these may help find/kill what is using port 2375
$ lsof -i :2375
$ kill <PID>
# configure docker to use the ssh tunnel
$ export DOCKER_HOST=:2375
You will probably want to unset DOCKER_HOST, if you plan to use docker-compose locally for testing (and you should).
$ unset DOCKER_HOST
Remember to
export DOCKER_HOST=:2375after setting up your ssh tunnel.
You can use the scripts in bin to switch between azure container service and local docker swarm.
$ docker-compose up
# if all went well, you should be able to visit your new docker swarm app hosted on azure at something like:
http://ti-acs-swarmagents.southcentralus.cloudapp.azure.com:8080/
# You can find the exact url using azure Resource Explorer
You may find it convenient to use ngrok to debug locally running services that require internet callbacks.
example config:
tunnels:
dev:
addr: 3001
proto: http
hostname: ngrok.transmute.industries
start the local host tunnel:
$ ngrok start dev