CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file. The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

• Changed project name from OpenAppStack to Stackspin
• Update:
  • Kube-prometheus-stack to helm chart version 18.0.12

0.7.0 - 2021-08-19

Removed

• BREAKING: The oas-secrets helm chart has been removed (#891) Because of this flux will remove your existing secrets when you change the tracking branch from v0.6 to v0.7. Follow our upgrade guide to prevent data loss!
• Unused flux notification-controller (#840)

Changed

• Docuementation overhaul (especially for installatation instructions)
• Increase cpu limits for eventrouter (#920)
• Increase mem limits for loki (#910)
• Retry taiko CI test until success
• Enable startupProbe for Nextcloud, remove initialDelaySeconds (#914)
• Reduce loki retention to 8 days because it uses too much disk space
• Update:
  • Taiko to 1.2.6
  • OAS Nextcloud chart to 0.3.2 (Nextcloud app to 20.0)
  • Dockerfile packages
• Pytest now tests cluster resources like deployments, statfulsets for readiness
• Optimize taiko chrome instances for parallel runs (#854)

Added

• Apps:
  • Wekan 5.41 (chart 5.42)
  • Nextcloud passwords 2021.7.23
  • Nextcloud Calendar 2.3.12
• Wekan, NC passwords and NC calendar to documentation

Fixed

• Helm chart values overrdides (#873)
• Prometheus CI test (#766)

0.6.0 - 2021-07-14

Added

• Option to use a container image mirror
• .sourceignore file for flux
• Eventrouter to aggregate k8s events in loki
• Fine-tune app resource limits
• Configure outgoing mail for alertmanager
• Code of conduct (CONTRIBUTING.md)
• Contribution docs (CONTRIBUTING.md)
• pre-commit config (.pre-commit-config.yaml)
• Add IP as SAN to k8s cluster API cert
• Add Lets encrypt staging certificate to cert store in test container

Changed

• Overhaul installation procedure
  • Seperate kubernetes and Openappstack installation
  • Secrets are now only managed on the cluster
  • settings.yml has been replaced with .flux.env and is saved as a secret in the cluster
  • Applications can be installed after first install of the cluster
  • Migrate to Flux 2
• Update documentation to reflect new installation procedure and flux2
• Update all helm charts and applications
• Flux 2
  • Increase all pull intervals to 1h
  • Change way of customizing helm charts
• Migrate from behave to taiko for end-to-end tests
• Update k3s and kubernetes version
  • Increase k3s event retention from 1h to 2d
• Loki:
  • Limit storage size
  • Increase chunk size, reduce chunk count
  • Increase retention time
  • Install loki + promtail seperatly, instead of loki-stack
• Use kube-prometheus-stack instead of prometheus + grafana
• Expose alertmanager and prometheus web interface by default
• Increase prometheus retention time
• CI
  • Only install and test apps on changes affecting an app
  • Improve pipeline speed and overall stability
  • Use new disk image template that has haveged pre-installed
  • Update Dockerfile base image to alpine 3.14
  • Test Onlyoffice Nextcloud integration
  • Allow invalid certs for OnlyOffice
  • Use 12GB (=6 cores) for master CI droplet
  • Disable the reuse of VPSs if there is already one

Removed

• remove old and now unused variables from compatibility-checks role