[aws-eks]: CDK uses integer for timeout; Helm expects duration

Wanted to test the new timeout parameter in the EKS Helm construct as I had problems deploying a large Helm chart like the prometheus-operator chart. But I just ran it and hit a problem: CDK translates the timeout to an integer value in seconds but the Helm version used (v3) expects a duration specification. So 900s instead of 900.

Reproduction Steps

        cluster.addChart('prometheus-operator', {
            chart: 'prometheus-operator',
            namespace: 'monitoring',
            release: 'prometheus-operator',
            repository: 'https://kubernetes-charts.storage.googleapis.com',
            timeout: Duration.minutes(15),
            wait: true

Error Log

[ERROR] Exception: b'Error: invalid argument "900" for "--timeout" flag: time: missing unit in duration 900\n' Traceback (most recent call last):   File "/var/task/index.py", line 16, in handler     return helm_handler(event, context)   File "/var/task/helm/__init__.py", line 49, in helm_handler     helm('upgrade', release, chart, repository, values_file, namespace, version, wait, timeout)   File "/var/task/helm/__init__.py", line 90, in helm     raise Exception(output)

Environment

• CLI Version :
• Framework Version:
• Node.js Version:
• OS :
• Language (Version):

Other

Not sure if this could be related to an upgrade @pahud is working on in aws-lambda-layer-kubectl?

---

This is 🐛 Bug Report

[eduardomourar]

I believe it was my mistake, so I will gladly fix it.

[eladb]

@eduardomourar thanks! let me know if you need assistance.