Files

Ilir Bekteshi f3a3e97618 Init mkdocs and gh workflow for publishing docs

Initialize mkdocs for serving documentation on GitHub pages
On push to ms-3 branch a workflow publishes the documentation to gh-pages which gets served by GitHub
The content is based on README.md

2020-03-11 21:32:34 +01:00

3.5 KiB

Raw Blame History

How qliksense cli works

At the initialization qliksense cli create few files in the director ~/.qliksene and it contains following files

.qliksense
├── config.yaml
├── contexts
│   └── qlik-default
│       └── qlik-default.yaml
└── ejson
    └── keys

qlik-default.yaml is a default CR has been created with some default values like this

apiVersion: qlik.com/v1
kind: Qliksense
metadata:
  name: qlik-default
spec:
  profile: docker-desktop
  secrets:
    qliksense:
    - name: mongoDbUri
      value: mongodb://qlik-default-mongodb:27017/qliksense?ssl=false
  rotateKeys: "yes"
  releaseName: qlik-default

The qliksense cli creates a default qliksense context (different from kubectl context) named qlik-default which will be the prefix for all kubernetes resources created by the cli under this context latter on. New context and configuration can be created by the cli.

$ qliksense config -h
do operations on/around CR

Usage:
  qliksense config [command]

Available Commands:
  apply         generate the patchs and apply manifests to k8s
  list-contexts retrieves the contexts and lists them
  set           configure a key value pair into the current context
  set-configs   set configurations into the qliksense context as key-value pairs
  set-context   Sets the context in which the Kubernetes cluster and resources live in
  set-secrets   set secrets configurations into the qliksense context as key-value pairs
  view          view the qliksense operator CR

Flags:
  -h, --help   help for config

Use "qliksense config [command] --help" for more information about a command.

qliksense cli works in two modes

with a git repo fork/clone of qliksense-k8s
without git repo

Without git repo

In this mode qliksense CLI download the specified version from qliksense-k8s and put it into folder ~/.qliksense/contexts/<context-name>/qlik-k8s.

The qliksense cli create a CR for the qliksense operator and all the config operations are peformed to edit the CR. So when qliksense install or qliksense config apply both generate patches in local file system (i.e ~/.qliksense/contexts/<context-name>/qlik-k8s) and install those manifests into the cluster and create a custom resoruce (CR) for the qliksene operator then the operator make association to the isntalled resoruces so that when qliksenes uninstall is performed the operator can delete all those kubernetes resources related to QSEoK for the current context.

With a git repo

User has to create fork or clone of qliksense-k8s and push it to their own git server. When user perform qliksense install or qliksene config apply the qliksense operator do these tasks

downloads the corresponding version of manifests from the user's git repo.
generate kustomize patches
install kubernetes resoruces
push those generated patches into a new branch in the provided git repo. so that user user can merge those patches into their master branch.
spinup a cornjob to monitor master branch. If user modifies anything in the master branch those changes will be applied into the cluster. This is a light weight git-ops model

This is how repo info is provided into the CR

qliksense config set git.repository="https://github.com/my-org/qliksense-k8s"

qliksense config set git.accessToken=blablalaala

3.5 KiB Raw Blame History

How qliksense cli works

Without git repo

With a git repo

3.5 KiB

Raw Blame History