Comment cost-saving changes in sandbox Helm Chart (#12424)

These are comments for 3 of the main changes that prevent lingering
containers from causing runaway costs in sandbox.
This commit is contained in:
Robert Fairburn 2023-06-21 14:35:48 -05:00 committed by GitHub
parent d70ae633c0
commit 2649073d91
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 8 additions and 0 deletions

View File

@ -10,8 +10,13 @@ metadata:
name: {{ .Values.fleetName }} name: {{ .Values.fleetName }}
namespace: {{ .Release.Namespace }} namespace: {{ .Release.Namespace }}
spec: spec:
# Forbiding concurrency prevents runaway costs on failing cronjobs stacking up
# see https://docs.google.com/document/d/1-4KmOlgfGEksNZnQo79a9nRLgM_i7ar2qovoZO3s_6g/edit.
concurrencyPolicy: Forbid concurrencyPolicy: Forbid
schedule: "{{ .Values.crons.vulnerabilities }}" schedule: "{{ .Values.crons.vulnerabilities }}"
# EKS Fargate keeps resources running to show the job history.
# This saves significantly on compute in AWS.
# https://docs.google.com/document/d/1-4KmOlgfGEksNZnQo79a9nRLgM_i7ar2qovoZO3s_6g/edit
successfulJobsHistoryLimit: 0 successfulJobsHistoryLimit: 0
jobTemplate: jobTemplate:
spec: spec:

View File

@ -10,6 +10,9 @@ metadata:
name: {{ .Values.fleetName }}-migration name: {{ .Values.fleetName }}-migration
namespace: {{ .Release.Namespace }} namespace: {{ .Release.Namespace }}
spec: spec:
# This will clean up the job to prevent excess costs when using
# EKS/Fargate. See
# https://docs.google.com/document/d/1-4KmOlgfGEksNZnQo79a9nRLgM_i7ar2qovoZO3s_6g/edit
ttlSecondsAfterFinished: 100 ttlSecondsAfterFinished: 100
template: template:
metadata: metadata: