CMG

Python Script to process text files and return status of instruments. Helm Chart Included

Tech

Pre-Req

Kubernetes Cluster Helm Chart

Ideally, some sort of ingress but port-forward will work for POC

Usage

cd chart/cmg
helm install cmg ./cmg-0.1.0.tgz
kubectl --namespace default port-forward service/cmg 8080:80

curl -X POST -F file=@data.txt http://localhost:8080/process

Results

{"hum-1":"Keep","hum-2":"Discard","temp-1":"Precise","temp-2":"Ultra Precise"}

Running Multiple Pods

❯ kubectl get pods
NAME                   READY   STATUS    RESTARTS   AGE
cmg-7f5d76ddb9-ml6cv   1/1     Running   0          19s
cmg-7f5d76ddb9-pmdq2   1/1     Running   0          19s

Things I wish I could change

Input file format is less than ideal.
Currently adding additional devices/rules requires editing the source code. Would be preferable if data ingestion could be seperate from business rules.
No CI/CD Process is in place. Would be nice to have time to configure Circle/Travis/Github Action to Spin up a KIND cluster, deploy the chart and verify it runs.
Helm Chart is not published.
No tests for python code. Since I knew I wouldnt get to a CI process, I didnt see any point in spending my 4 hours on the tests.
No Infra code. I only have access to AWS, and I didnt want to pay for the spend to spin up physical resources via TF.

Adding additional Devices/Rules

Extend this function

# Each Devices needs a rule to return its status
def rule_validation(device_type, name, metrics, ref_temp, ref_humidity):
    """Determine Rules for Device types"""
    metrics = list(map(float, metrics))
    if device_type == "thermometer":
        return check_thermometer(metrics, ref_temp)
        
    elif device_type == "humidity":
        return check_humidity(metrics, ref_humidity)

Then add a custom validation function for that device type.

def check_something(metrics, ref_temp):
    return "Pass"