Commit c27ad3b3 authored by garciadeblas's avatar garciadeblas
Browse files

Merge branch 'revert-47df6c46' into 'master' 

Revert "K8s_01-Robot tests packages"

See merge request !75
parents 71472f28 385f0acc

k8s_juju_ns/README.md

deleted100644 → 0
+0 −159
Original line number Diff line number Diff line 
# Onboarding and instantiation



```bash

osm nfpkg-create k8s_base_vnf.tar.gz

osm nfpkg-create k8s_juju_vnf.tar.gz

osm nspkg-create k8s_base_ns.tar.gz

osm ns-create --ns_name k8s --nsd_name k8s_base --vim_account ost9-canonical-fortville --config '{vld: [ {name: mgmtnet, vim-network-name: mgmt} ] }'

```



This deploys a NS (k8s\_base\_ns) with one VNF that runs Juju (k8s\_juju\_vnf) and 4 VNFs (k8s\_base\_vnf) that run the K8s cluster (1 master, 3 workers).



VMs will be called: 



- JUJU

- NODE1

- NODE2

- NODE3

- NODE4



# Install and bootstrap of Juju in the VM where juju runs (JUJU)



## Install juju



```bash

sudo snap install juju --classic --channel=2.6/stable

ssh-keygen

```



## Add the cloud: manual provider type



```bash

# Copy the ssh key locally

ssh-copy-id ubuntu@<IP_JUJU>

juju add-cloud

    manual

    k8s

    ubuntu@<IP_JUJU>

```



## Create the controller



```bash

ssh-copy-id -i /home/ubuntu/.local/share/juju/ssh/juju_id_rsa ubuntu@<IP_JUJU>

#The same as: cat /home/ubuntu/.local/share/juju/ssh/juju_id_rsa.pub >> .ssh/authorized_keys

juju bootstrap k8s jujuk8s

```



This creates a controller locally in the machine



Juju status can be checked with:



```

juju status

```



# Add the VMs to juju (manual provider mode)



First from JUJU you need to copy the controller SSH key to all nodes of the cluster (master and workers)



```bash

#cat /home/ubuntu/.local/share/juju/ssh/juju_id_rsa.pub

#vi .ssh/authorized_keys

ssh-copy-id -i /home/ubuntu/.local/share/juju/ssh/juju_id_rsa ubuntu@<IP_NODE1>

ssh-copy-id -i /home/ubuntu/.local/share/juju/ssh/juju_id_rsa ubuntu@<IP_NODE2>

ssh-copy-id -i /home/ubuntu/.local/share/juju/ssh/juju_id_rsa ubuntu@<IP_NODE3>

ssh-copy-id -i /home/ubuntu/.local/share/juju/ssh/juju_id_rsa ubuntu@<IP_NODE4>

```



After that add the machines to juju controller mannually in JUJU: 



```bash

juju add-machine ssh:ubuntu@<IP_NODE1> --debug

juju add-machine ssh:ubuntu@<IP_NODE2> --debug

juju add-machine ssh:ubuntu@<IP_NODE3> --debug

juju add-machine ssh:ubuntu@<IP_NODE4> --debug

```



Next we look for the indentifiers of the machines and take notes of it.



```bash

juju machines

```



If they are the first machines we add, then the ids will be 0, 1, 2, 3.



# Deploy the K8s cluster with the bundle



The original bundle cannot be used as it is because it assumes that it is going to be deployed in a cloud without specifying the machines.



```bash

#juju deploy charmed-kubernetes

#wget https://api.jujucharms.com/charmstore/v5/charmed-kubernetes/archive/bundle.yaml

#vi bundle.yaml

```



The bundle has been edited to specify which application runs in which machine.



We copy the bundle from osm-descriptors repo to JUJU:



```bash

scp osm-descriptors/tid/k8s/k8s_base_ns/bundle/bundle.yaml ubuntu@<IP_JUJU>

```



Deploy the bundle and check the result:



```bash

juju deploy ./bundle.yaml --map-machines=existing,0=0,1=1,2=2,3=3

#juju deploy ./bundle.yaml --map-machines=existing,0=30,1=31,2=32,3=33

juju status

```



# Interacting with the Kubernetes cluster



In JUJU machine:



```bash

mkdir -p ~/.kube

juju scp kubernetes-master/0:config ~/.kube/config

sudo snap install kubectl --classic

kubectl cluster-info

```



# To access the K8s dashboard 



In JUJU machine:



```bash

juju config kubernetes-master enable-dashboard-addons=true

```



The dashboard could be accesible running this kubectl command:



```bash

kubectl proxy --address <IP_JUJU>

```



And now a SSH tunnel could be done from the PC to JUJU, and from the web browser access to the dashboard:



<http://localhost:8001/api/v1/namespaces/kube-system/services/https:kubernetes-dashboard:/proxy/>



# Install and init Helm



## Install



```bash

sudo snap install helm --classic

```



## Init



```bash

# Optional. Create account for tiller

# kubectl --namespace kube-system create serviceaccount tiller

# kubectl create clusterrolebinding tiller-cluster-rule --clusterrole=cluster-admin --serviceaccount=kube-system:tiller

helm init

# helm init --service-account tiller --wait

# kubectl --namespace kube-system patch deploy tiller-deploy -p '{"spec":{"template":{"spec":{"serviceAccount":"tiller"}}}}'

kubectl get all --all-namespaces

```

k8s_juju_ns/bundle/bundle.yaml

deleted100644 → 0
+0 −133
Original line number Diff line number Diff line 
description: A highly-available, production-grade Kubernetes cluster.

series: bionic

machines:

  "0":

  "1":

  "2":

  "3":

services:

  containerd:

    annotations:

      gui-x: '475'

      gui-y: '800'

    charm: cs:~containers/containerd-20

    resources: {}

  easyrsa:

    annotations:

      gui-x: '90'

      gui-y: '420'

    charm: cs:~containers/easyrsa-270

    constraints: root-disk=8G

    num_units: 1

    resources:

      easyrsa: 5

    to:

    - "0"

  etcd:

    annotations:

      gui-x: '800'

      gui-y: '420'

    charm: cs:~containers/etcd-449

    constraints: root-disk=8G

    num_units: 3

    options:

      channel: 3.2/stable

    resources:

      core: 0

      etcd: 3

      snapshot: 0

    to:

    - "1"

    - "2"

    - "3"

  flannel:

    annotations:

      gui-x: '475'

      gui-y: '605'

    charm: cs:~containers/flannel-438

    resources:

      flannel-amd64: 394

      flannel-arm64: 390

      flannel-s390x: 377

  kubeapi-load-balancer:

    annotations:

      gui-x: '450'

      gui-y: '250'

    charm: cs:~containers/kubeapi-load-balancer-664

    constraints: root-disk=8G

    expose: true

    num_units: 1

    resources: {}

    to:

    - "0"

  kubernetes-master:

    annotations:

      gui-x: '800'

      gui-y: '850'

    charm: cs:~containers/kubernetes-master-724

    constraints: cores=2 mem=4G root-disk=16G

    num_units: 2

    options:

      channel: 1.15/stable

    resources:

      cdk-addons: 0

      core: 0

      kube-apiserver: 0

      kube-controller-manager: 0

      kube-proxy: 0

      kube-scheduler: 0

      kubectl: 0

    to:

    - "1"

    - "2"

  kubernetes-worker:

    annotations:

      gui-x: '90'

      gui-y: '850'

    charm: cs:~containers/kubernetes-worker-571

    constraints: cores=4 mem=4G root-disk=16G

    expose: true

    num_units: 3

    options:

      channel: 1.15/stable

    resources:

      cni-amd64: 392

      cni-arm64: 383

      cni-s390x: 395

      core: 0

      kube-proxy: 0

      kubectl: 0

      kubelet: 0

    to:

    - "1"

    - "2"

    - "3"

relations:

- - kubernetes-master:kube-api-endpoint

  - kubeapi-load-balancer:apiserver

- - kubernetes-master:loadbalancer

  - kubeapi-load-balancer:loadbalancer

- - kubernetes-master:kube-control

  - kubernetes-worker:kube-control

- - kubernetes-master:certificates

  - easyrsa:client

- - etcd:certificates

  - easyrsa:client

- - kubernetes-master:etcd

  - etcd:db

- - kubernetes-worker:certificates

  - easyrsa:client

- - kubernetes-worker:kube-api-endpoint

  - kubeapi-load-balancer:website

- - kubeapi-load-balancer:certificates

  - easyrsa:client

- - flannel:etcd

  - etcd:db

- - flannel:cni

  - kubernetes-master:cni

- - flannel:cni

  - kubernetes-worker:cni

- - containerd:containerd

  - kubernetes-worker:container-runtime

- - containerd:containerd

  - kubernetes-master:container-runtime

k8s_juju_ns/icons/osm.png

deleted100644 → 0
−54.6 KiB
Loading image diff...

k8s_juju_ns/k8s_juju_nsd.yaml

deleted100644 → 0
+0 −43
Original line number Diff line number Diff line 
nsd-catalog:

    nsd:

    -   id: k8s_juju

        name: k8s_juju

        short-name: k8s_juju

        description: NS consisting of a 4 k8s_jujumachine VNFs and 1 k8s_jujucontroller VNF connected to mgmt network

        vendor: OSM

        version: '1.0'

        logo: osm.png

        constituent-vnfd:

        -   member-vnf-index: k8s_vnf1

            vnfd-id-ref: k8s_jujumachine_vnf

        -   member-vnf-index: k8s_vnf2

            vnfd-id-ref: k8s_jujumachine_vnf

        -   member-vnf-index: k8s_vnf3

            vnfd-id-ref: k8s_jujumachine_vnf

        -   member-vnf-index: k8s_vnf4

            vnfd-id-ref: k8s_jujumachine_vnf

        -   member-vnf-index: k8s_juju

            vnfd-id-ref: k8s_jujucontroller_vnf

        vld:

        -   id: mgmtnet

            name: mgmtnet

            type: ELAN

            mgmt-network: 'true'

            vim-network-name: mgmt

            vnfd-connection-point-ref:

            -   member-vnf-index-ref: k8s_vnf1

                vnfd-id-ref: k8s_jujumachine_vnf

                vnfd-connection-point-ref: mgmt

            -   member-vnf-index-ref: k8s_vnf2

                vnfd-id-ref: k8s_jujumachine_vnf

                vnfd-connection-point-ref: mgmt

            -   member-vnf-index-ref: k8s_vnf3

                vnfd-id-ref: k8s_jujumachine_vnf

                vnfd-connection-point-ref: mgmt

            -   member-vnf-index-ref: k8s_vnf4

                vnfd-id-ref: k8s_jujumachine_vnf

                vnfd-connection-point-ref: mgmt

            -   member-vnf-index-ref: k8s_juju

                vnfd-id-ref: k8s_jujucontroller_vnf

                vnfd-connection-point-ref: mgmt

k8s_jujucontroller_vnf/README

deleted100644 → 0
+0 −3
Original line number Diff line number Diff line 
This is a VNF with a single small VDU intended to run a juju controller

that will deploy a K8s cluster using a juju bundle on a set of machines.