https://github.com/gavindsouza/robocorp-workers-bench
Simplified Robocorp Worker Management Interface
https://github.com/gavindsouza/robocorp-workers-bench
robocorp worker
Last synced: 3 months ago
JSON representation
Simplified Robocorp Worker Management Interface
- Host: GitHub
- URL: https://github.com/gavindsouza/robocorp-workers-bench
- Owner: gavindsouza
- License: mit
- Created: 2023-04-27T10:28:32.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-08-15T13:43:16.000Z (almost 2 years ago)
- Last Synced: 2025-01-15T14:18:20.944Z (4 months ago)
- Topics: robocorp, worker
- Language: Python
- Homepage:
- Size: 24.4 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
### Simplified Robocorp Setup
Run `./worker-up.py` for available CLI options & help.
```bash
usage: Robocorp Worker Wrapper [-h] [--name NAME] [--token TOKEN] [--build] [--dockerfile-location DOCKERFILE_LOCATION] [--image-name IMAGE_NAME]
[--image-version IMAGE_VERSION]
Manage Robocorp self-hosted workers
options:
-h, --help show this help message and exit
--name NAME, -n NAME
--token TOKEN, -t TOKEN
--build, -b Build the worker image
--dockerfile-location DOCKERFILE_LOCATION, -d DOCKERFILE_LOCATION
--image-name IMAGE_NAME, -i IMAGE_NAME
--image-version IMAGE_VERSION, -v IMAGE_VERSION
Simplified Robocorp Worker Management Interface
```
#### Steps to setup a Cluster of self-hosted workers
Steps to build & spin up a self-hosted worker:
1. Generate a `Dockerfile` that outlines the worker environment
I prefer using the dynamic workers for quickly spinning up workers that I can discard later.
```bash
cp ./base/dynamic/Dockerfile Dockerfile
```
2. Setup appropriate conda.yaml that outlines robot runtime context
A conda file that resembles the environment your robots need. This ensures a suitable environment is generated while building the image to avoid any slow downs during executing robot processes.
```bash
cp {your-robot-conda} conda.yaml
```
3. Build a worker image
Build the common image which will be used for spinning up new workers.
```bash
./worker-up.py --build -i rc_dynamic -v 1.0
```
4. Spin up a worker using a single or group token
Each worker started translates to a container spun up using the image generated in the previous step.
```bash
./worker-up.py -i rc_dynamic -v 1.0 -n {worker_name} -t {token}
```
#### Gavin's Workflow
1. Provision a new server & SSH to it
1. `apt update && apt upgrade -y && apt install git`
1. `curl https://get.docker.com/ | sh`
1. `git clone https://github.com/gavindsouza/robocorp-workers-bench robocorp`
1. `cd robocorp`
1. `cp ./base/dynamic/Dockerfile .`
1. `cp ./base/conda.yaml .`
1. `./worker-up.py --build`
1. `export ROBOCORP_GROUP_TOKEN=${NEW_TOKEN_GENERATED_IN_CONTROL_ROOM}`
1. ```for i in `seq 1 N`; do ./worker-up.py -n "a-$i" -t $ROBOCORP_GROUP_TOKEN; done```
```bash
# for copy pastin' - update token & worker count values (first & last lines)
export ROBOCORP_GROUP_TOKEN=${NEW_TOKEN_GENERATED_IN_CONTROL_ROOM}
apt update && apt upgrade -y && apt install git
curl https://get.docker.com/ | sh
git clone https://github.com/gavindsouza/robocorp-workers-bench robocorp
cd robocorp
cp ./base/dynamic/Dockerfile .
cp ./base/conda.yaml .
./worker-up.py --build
for i in `seq 1 N`; do ./worker-up.py -n "a-$i" -t $ROBOCORP_GROUP_TOKEN; done
```
#### Notes
Copy the `conda.yaml` from the targetted robot to this folder. It will be used in the container image building process.
Added the `--no-cache` during the image building because we don't want to cache the `rcc` environment creation. Alternatively, we can add the `ADD` directive to the pypi & github URLs to do so automatically.