Skip to content
This repository has been archived by the owner on Mar 12, 2021. It is now read-only.

Latest commit

 

History

History
445 lines (315 loc) · 14.3 KB

README.md

File metadata and controls

445 lines (315 loc) · 14.3 KB
Raw

Vamp Kubist command line client

CircleCI Coverage Status

Vamp Kubist Command Line Client is a command line client written in golang and allows a user interface to the Vamp Kubist API. It also supports common functions that are useful to regular users.

Command Line client requires a running Vamp Kubist service to function.

Important Note:

Please check How to Setup and Use Vamp Kubist on common documentation repository:

https://github.com/magneticio/vampkubistdocs

This document only shows only the basic usage of the client.

development

if you have golang installed, it is recommended to git clone it to $GOPATH/src/github.com/magneticio/vampkubistcli This is a requirement for docker builder to work.

It is recommended to read and follow golang setup for a development environment setup: https://golang.org/doc/install

build

If you get errors about missing libraries while building, run:

go get

for docker build:

./build.sh

for local build:

./build.sh local

binaries will be under bin directory

installation

If you have binaries built locally: For mac run:

./bin/vamp-darwin-amd64 --help

If you have downloaded the binary directly, Just copy the binary for you platform to the user binaries folder for general usage, for MacOS:

cp vamp-darwin-amd64 /usr/local/bin/vamp
chmod +x /usr/local/bin/vamp

If you don't have anything yet and automatically download an install follow commands for your platform:

This installation may not work since this is a private repository. Manual installation is recommended.

Easy install for MacOS / Linux:

KUBIST_CLI_VERSION=$(curl -s https://api.github.com/repos/magneticio/vampkubistcli/releases/latest | grep '"tag_name":' | sed -E 's/.*"([^"]+)".*/\1/') && \
    base=https://github.com/magneticio/vampkubistcli/releases/download/$KUBIST_CLI_VERSION && \
    curl -fsSL  $base/vamp-$(uname -s)-$(uname -m) > /usr/local/bin/vamp && \
    chmod +x /usr/local/bin/vamp

TODO: add installation for other platforms

For general users it is recommended to download the binary for your platform. Latest release can be found here: https://github.com/magneticio/vamp/releases/latest

Run

This documentation is prepared for explaining the general usage of command line client. This Readme doesn't explain every path. Check docs repository for more: https://github.com/magneticio/vampkubistdocs

Check the version of the client with:

vamp version

It is possible to get all commands and flags by running help:

vamp help

This documentation is for version 0.0.13 and above

First you need to login to the vamp application: You will need

  • the url of your cluster eg.: http://1.2.3.4:8888
  • username eg.: username
  • password eg.: password

Please ask your provider if you don't have these information

If you have installed your vamp into your cluster, these information will be given to you during the installation. If you are using the SaaS version, this information will be provided by the Vamp.io

Basic authorisation:

vamp login --url http://1.2.3.4:8888 --user username --password password

It is also possible to login with access token, refresh token and pass certificate for secure communication.

For managing all the resources, there is an umbrella concept called project. Project name should be unique inside a vamp application. By starting a new project, users can bundle all the resources under a project.

It is recommend to start by creating a Project.

Please list projects first and choose a new projec name:

vamp list project

Please set a project name that is not listed above

PROJECT_NAME=myproject

This command will let you create an empty project

vamp create project $PROJECT_NAME --init

"--init" flag is a shorthand for creating empty objects

Client allows you to pass specifications as json, yaml and from file or from string. This command reads the input as json and passes the empty json object as configuration. Project doesn't have any mandatory fields, so an empty json is still valid.

You can delete the project just created with:

vamp delete project $PROJECT_NAME

Please download resources folder from the repo to continue rest of the examples. Assuming resources folder exists in your current workspace;

Let's create a project from a yaml file currently exists in the resources folder.

contents of ./resources/project.yaml

metadata:
  key1: value1
  key2: value2
vamp create project $PROJECT_NAME -f ./resources/project.yaml

If you don't have files locally you can also load them from a remote location:

Resources for this readme is also located at https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1

vamp create project $PROJECT_NAME -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/project.yaml

Rest of the README will use remote resources.

You can check the details of the created project with get method:

vamp get project $PROJECT_NAME

By default it will print yaml representation of project For JSON user can specify output, eg.:

vamp get project $PROJECT_NAME -o json

Let's update the project configuration with an updated file.

contents of project_update.yaml

metadata:
  key1: value1
  key2: value2_new
vamp update project $PROJECT_NAME -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/project_update.yaml

Run get project again to see the changes:

vamp get project $PROJECT_NAME

You will see the second key in the metadata changed to "key2: value2_new"

This concept applies to every object in vamp system as long as you have access rights as a user.

Vamp has the concept of Project, Cluster, Virtual Cluster in the order of hierarchy.

If you have used kubernetes, Cluster represents configurations related to an actual cluster and Virtual Cluster is configurations bound to a namespace.

You are expected to work in the same project so you can set it as default by running set command:

vamp config set -p $PROJECT_NAME

It is possible to get current configuration with following command:

vamp config get

You can create a user with the following command:

contents of user1.yaml

userName: user1
password: pass1
vamp create user user1 -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/user1.yaml

And also grant admin role to user1 in your project:

vamp grant --user user1 --role admin -p $PROJECT_NAME

You can also create additional roles for simplicity, admin role is used.

To revoke the same role run:

vamp revoke --user user1 --role admin -p $PROJECT_NAME

Now login as the user1:

vamp login --user user1 --password pass1

It can be seen that while using the same cluster it is not needed to pass the url or certificate information again.

Now, you can bootstrap your cluster with vamp. Please make sure you have installed kubectl and authenticated to connect to the cluster you want to be managed by vamp. Command line client will set up a service account user in your cluster and set up credentials to connect to your cluster in vamp.

For this example, it is recommended that you have a cluster of at least 5 nodes, 2 CPU and 8 GB Ram per node. Otherwise you can have pending pods and initialisation will not finish.

vamp bootstrap cluster mycluster

A simple cluster configuration requires;

  • url
  • cacertdata
  • serviceaccount_token

You can check if your cluster is properly configured by running a get to the cluster.

vamp get cluster mycluster

with kubectl you can check the namespaces of vamp-system and istio-system and logging is created.

Vamp will run a job in vamp-system namespace to make sure that everything is properly installed and continue running this job until it is finished. Make sure that you have enough resources and there are no pending pods or pods in Crash Loop.

If you have watch command installed, I recommended running following watch command to see instalment in action:

watch kubectl get pods --all-namespaces

Run kubectl command to see if there are pending pods, which is a common issue when there is not enough resources:

kubectl get pods --all-namespaces | grep Pending

If there are pending pods after some a few minutes, it is recommended to diagnose the issue, and if it is a resource issue, scale up the cluster.

Again while working on the same cluster it is recommended to set it as default by:

vamp config set -c mycluster

Now it is time to deploy an example application with one version:

kubectl apply -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/demo-application-version1.yaml

This will create a namespace called vamp-demo and deploy two deployments. There are two ways of importing a namespace to vamp

  • label the namespace with "vamp-managed: enabled"
  • Create a virtual cluster through vamp

NOTE: Vamp Kubist does not create the namespace, it should exist before the creation. If the namespace does not exist, creation will not fail immediately, it will search for namespace and it will delete itself if the namespace does not exist. It is a combination of behaviours that required to scan for namespaces and not managing them.

vamp create virtual_cluster vamp-demo -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/virtualcluster.yaml

This will re-label the namespace with required settings if the namespace exits. It will not create the namespace, as it is expected to be created by a deployment pipeline.

kubectl get ns vamp-demo -o yaml

set virtual cluster for ease of use:

vamp config set -r vamp-demo

NOTE: In all networking structures (gateway, destination, service, etc.), ports should be defined explicitly. All ports are closed until they are defined to be opened.

To expose the application to outside, you will need a gateway:

vamp create gateway shop-gateway -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/gateway.yaml

Create a destination

vamp create destination shop-destination -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/destination.yaml

Final step is creating the VampService. An http load balanced VampService requires the hostnames. In this example hostname is the IP address, IP address of the gateway is easy to get with get command:

vamp get gateway shop-gateway

IP is under status > ip,

It is easier to get with a grep command:

vamp get gateway shop-gateway | grep ip

Or you can use jsonpath parameter to filter the ip, this is an experimental feature and only supports string parameters:

vamp get gateway shop-gateway -o=json --jsonpath '$.status.ip'

You can also wait until it is available:

vamp get gateway shop-gateway -o=json --jsonpath '$.status.ip' --wait

To set the IP address to a bash variable run: DON'T SKIP THIS STEP

GATEWAY_IP=$(vamp get gateway shop-gateway -o=json --jsonpath '$.status.ip' --wait)
echo $GATEWAY_IP

This command waits until IP is ready and set it to a bash parameter.

You need so see an IP address which is the IP address of the public facing load balancer.

I will reference this ip as GATEWAY_IP in the rest of the documentation.

Please note that this may take some time depending on the provider.

You need to copy paste the IP address of the gateway in the vamp service configuration under the hosts as an array. Since this is harder to do in remote resources, there is an easier way to add hosts on the fly with a host parameter.

Create a Vamp Service with 100% traffic on version1

vamp create vamp_service shop-vamp-service -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/vampservice_template.yaml --host $GATEWAY_IP

Check link generated with:

echo http://$GATEWAY_IP

Now your application is released to public, copy paste the ip of your gateway to your favourite browser (http://$GATEWAY_IP). Now you should be able to see the blue e-commerce page. This is the first version.

Let's deploy the second version of the e-commerce website with kubectl

kubectl apply -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/demo-application-version2.yaml

Now with this canary release, the second version will be released in time based manner, If there is an error in the responses, it will automatically roll back and retry until successful.

You can create a canary release with configuration file or you can use the release command.

Create with configuration:

vamp create canary_release shop-release -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/canaryrelease.yaml

Release command:

vamp release shop-vamp-service --destination shop-destination --subset subset2 -l version=version2

Check your browser and refresh frequently to see the second version is available.

You can also check the percentage changes with:

vamp get vamp_service shop-vamp-service

It will take some time to release totally and you can not see the first version anymore.

But now you decided, a url based access to these version are more useful for you, then you can set up conditional routes.

Before going forward delete current canary release

vamp list canary_release
vamp delete canary_release shop-vamp-service-shop-destination-subset2

Update the hosts field the IP address of gateway in ./resources/conditionalvampservice.yaml then update the vamp service:

vamp update vamp_service shop-vamp-service -f https://raw.githubusercontent.com/magneticio/demo-resources/master/vamplamiacliv1/conditionalvampservice_template.yaml --host $GATEWAY_IP

Now you can see that first version is available under url: http://$GATEWAY_IP/v1 and second version is under http://$GATEWAY_IP/v2

Check links generated with:

echo http://$GATEWAY_IP/v1
echo http://$GATEWAY_IP/v2