StackStorm in Docker containers

The default container configuration is as follows:

• stackstorm (st2 + st2web + st2mistral)
• mongo
• rabbitmq
• postgres
• redis

The mongo, rabbitmq, postgres and redis containers store their data on persistent storage. Additionally, the stackstorm container persists the contents of /var/log. If you do not wish to persist this data, then remove the appropriate entries from docker-compose.yml.

Usage

We use Version 3 of the compose file format, so if you want to run docker-compose, you'll need to ensure you're running Docker Engine release 1.13.0+.

First, execute

make env

to create the environment files used by docker-compose. You may want to change the values of the variables as necessary, but the defaults should be okay if you are not using any off-cluster services (e.g. mongo, redis, postgres, rabbitmq).

Below is the complete list of available options that can be used to customize your container.

Parameter	Description
ST2_USER	StackStorm account username
ST2_PASSWORD	StackStorm account password
MONGO_HOST	MongoDB server hostname
MONGO_PORT	MongoDB server port (typically 27017)
MONGO_DB	(Optional) MongoDB dbname (will use st2 if not specified)
MONGO_USER	(Optional) MongoDB username (will connect without credentials if this and MONGO_PASS are not specified)
MONGO_PASS	(Optional) MongoDB password
RABBITMQ_HOST	RabbitMQ server hostname
RABBITMQ_PORT	RabbitMQ server port (typically 5672)
RABBITMQ_DEFAULT_USER	RabbitMQ username
RABBITMQ_DEFAULT_PASS	RabbitMQ password
POSTGRES_HOST	PostgreSQL server hostname
POSTGRES_PORT	PostgreSQL server port (typically 5432)
POSTGRES_DB	PostgreSQL database
POSTGRES_USER	PostgreSQL username
POSTGRES_PASSWORD	PostgreSQL password
REDIS_HOST	Redis server hostname
REDIS_PORT	Redis server port
REDIS_PASSWORD	(Optional) Redis password

Second, start the docker environment.

docker-compose up -d

This will pull the required images from docker hub, and then start them.

However, if you find need to modify the stackstorm image, you will need to build it. Run:

REPO=stable
docker build --build-arg ST2_REPO=${REPO} stackstorm/stackstorm:${REPO}

where REPO is one of 'stable', 'unstable', 'staging-stable', 'staging-unstable'. Otherwise, the following docker-compose command will download the specified image from docker hub.

To stop the docker environment, run:

docker-compose down

Running custom shell scripts on boot

This container supports running arbitrary shell scripts on container boot. Any *.sh file located under /entrypoint.d directory will be executed inside the container just before starting stackstorm services.

For example, if you want to modify /etc/st2/st2.conf to set system_packs_base_path parameter, create modify-st2-config.sh with the follwing content:

/bin/bash
crudini --set /etc/st2/st2.conf content system_packs_base_path /opt/stackstorm/custom_packs

Then bind mount it to /entrypoint.d/modify-st2-config.sh

• via docker run

  docker run -it -d --privileged \
    -v /path/to/modify-st2-config.sh:/entrypoint.d/modify-st2-config.sh \
    stackstorm/stackstorm:latest
  

• via docker-compose.yml

  services:
    stackstorm:
      image: stackstorm/stackstorm:${TAG:-latest}
       : (snip)
      volumes:
        - /path/to/modify-st2-config.sh:/entrypoint.d/modify-st2-config.sh
  

The above example shows just modifying st2 config but basically there is no limitation so you can do almost anything.

You can also bind mount a specific directory to /entrypoint.d then place scripts as much as you want. All of them will be executed as long as the file name ends with *.sh.

Note: scripts will be executed in alphabetical order of the file name.

Adding a simple action

We will add a simple action that runs a local shell command. Run the following from your docker host.

mkdir -p packs.dev/examples/actions
cp examples/actions/hello.yaml packs.dev/examples/actions

Get a bash shell in the stackstorm container:

docker exec -it stackstorm /bin/bash

Load the new action into StackStorm. Whenever you change the yaml file, you need to run st2ctl reload. Within the container, run the following:

root@aff39eda0bdd:/# st2ctl reload --register-all

... output trimmed ...

Now, let's run the action:

root@aff39eda0bdd:/# st2 run examples.hello
.
id: 58f67dbf33a99300bdc4d618
status: succeeded
parameters: None
result:
  failed: false
  return_code: 0
  stderr: ''
  stdout: Hello human!
  succeeded: true

The action takes a single parameter name, which as we can see above, defaults to 'human' if name is not specified. If we specify a value for name, then as expected, the value is found in result.stdout.

root@aff39eda0bdd:/# st2 run examples.hello name=Stanley
.
id: 58f67dc533a99300bdc4d61b
status: succeeded
parameters:
  name: Stanley
result:
  failed: false
  return_code: 0
  stderr: ''
  stdout: Hello Stanley!
  succeeded: true

Congratulations, you have created your first simple action!

A Slight Variation: Concurrency

If you want to take advantage of concurrency, use a slight variation on the above. On the host, run:

mkdir -p packs.dev/examples/policies
cp examples/actions/hello-concurrency.yaml packs.dev/examples/actions
cp examples/policies/hello-concurrency.yaml packs.dev/examples/policies

Inside the stackstorm container, run:

st2ctl reload --register-all

Open two terminals to the stackstorm container. In the first, type (but don't execute):

st2 run examples.hello-concurrency name=1

In the second, type:

st2 run examples.hello-concurrency name=2

Now, execute the command in the first terminal, wait 5 seconds and then execute the command in the second terminal. After a second or so, you should see the following in the second terminal:

root@258b11849aa7:/# st2 run examples.hello-concurrency name=2
.
id: 590cec228964ad01567f61e3
status: delayed
parameters:
  name: 2
result: None

If you run st2 execution list before 10 seconds have elapsed, the status of the second action should be "delayed". Between 10 and 20 seconds, the status of the second action should be "running". After 20 seconds, the status of the second action should be "succeeded".

+--------------------------+----------------------------+--------------+-------------------------+-------------------------------+-------------------------------+
| id                       | action.ref                 | context.user | status                  | start_timestamp               | end_timestamp                 |
+--------------------------+----------------------------+--------------+-------------------------+-------------------------------+-------------------------------+
| 590cec068964ad01567f61dd | examples.hello-concurrency | st2admin     | succeeded (10s elapsed) | Fri, 05 May 2017 21:17:58 UTC | Fri, 05 May 2017 21:18:08 UTC |
| 590cec1f8964ad01567f61e0 | examples.hello-concurrency | st2admin     | succeeded (10s elapsed) | Fri, 05 May 2017 21:18:23 UTC | Fri, 05 May 2017 21:18:33 UTC |
| 590cec228964ad01567f61e3 | examples.hello-concurrency | st2admin     | succeeded (17s elapsed) | Fri, 05 May 2017 21:18:26 UTC | Fri, 05 May 2017 21:18:43 UTC |
+--------------------------+----------------------------+--------------+-------------------------+-------------------------------+-------------------------------+

Adding a rule

To perform a very basic end-to-end test of StackStorm, let's create a simple rule. Run the following from your docker host.

mkdir packs.dev/examples/rules
cp examples/rules/monitor_file.yaml packs.dev/examples/rules

We need to tell the FileWatchSensor to watch /tmp/date.log, enable the linux.FileWatchSensor and then call st2ctl reload.

Use docker exec to connect to the stackstorm container:

docker exec -it stackstorm /bin/bash

Within the container, run the following:

echo "    - /tmp/date.log" >> /opt/stackstorm/packs/linux/config.yaml
st2ctl reload
st2 sensor enable linux.FileWatchSensor

When we append to /tmp/date.log, the sensor will inject a trigger that matches the criteria. The linux.file_touch action is called, creating /tmp/touch.log.

Now let's append a line to the file in the container.

echo "hi" >> /tmp/date.log

The file /tmp/touch.log should exist with a recent timestamp. Congratulations, you have created your first rule!

Adding a python action

As an example of how to create a new action, let's add a new action called echo_action.

First, on the host, we create the metadata file ./packs.dev/examples/actions/my_echo_action.yaml:

---
name: "echo_action"
runner_type: "python-script"
description: "Print message to standard output."
enabled: true
entry_point: "my_echo_action.py"
parameters:
  message:
    type: "string"
    description: "Message to print."
    required: true
    position: 0

Then, add the action script at ./packs.dev/examples/actions/my_echo_action.py.

import sys

from st2actions.runners.pythonrunner import Action

class MyEchoAction(Action):
  def run(self, message):
    print(message)

    if message == 'working':
      return (True, message)
    return (False, message)

When you rename, or create a new action, you must run st2ctl reload inside the st2 container. Next, to initialize the virtualenv, run:

  st2 run packs.setup_virtualenv packs=examples

Then you can run your action using the following:

  st2 run examples.echo_action message=working

You should see output similar to:

.
id: 58c0abcff4aa45009f42dca3
status: succeeded
parameters:
  message: working
result:
  exit_code: 0
  result: working
  stderr: ''
  stdout: 'working

    '

Congratulations! You have successfully added your first action!

Adding a simple mistral workflow

To add a simple mistral workflow, run the following from your docker host.

mkdir -p packs.dev/examples/actions/workflows
cp -R examples/actions/mistral-basic.yaml packs.dev/examples/actions/mistral-basic.yaml
cp -R examples/actions/workflows/mistral-basic.yaml packs.dev/examples/actions/workflows/mistral-basic.yaml

Use docker exec to connect to the stackstorm container:

docker exec -it stackstorm /bin/bash

Within the container, run the following:

st2 action create /opt/stackstorm/packs.dev/examples/actions/mistral-basic.yaml
st2 run examples.mistral-basic cmd=date -a

The st2 run command should complete successfully. Please see mistral documentation for more details about this basic workflow.

Congratulations, you have created your first mistral workflow!