Cloud实战:Manage and Manipulate OpenStack with CLI, SDK and API

标签：openstack cli sdk api aws hybrid cloud

 Cloud IN ACTION: Manage and Manipulate OpenStack with CLI, SDK and API

  

 薛国锋    xueguofeng2011@gmail.com

Today we are going to experience the three ways to manage and manipulate OpenStack –CLI, SDK and API; and then develop a simple Hybrid Cloud application with Python to exchange data between Amazon Web Services and OpenStack.

 

API, SDK and CLI

 

There are three ways you canuse Openstack:

 

RESTful API, you can send the requests directly using HTTP GET or POST via a Web Browser, your program code or using cURL to the service endpoints (APIGateway) of OpenStack and issue your commands. Using the OpenStack APIs, you can launch server instances, create images, assign metadata to instances and images, create storage containers and objects, and complete other actions in the OpenStack-based Cloud.

 

SDK, aka the language-level API. Making the API calls directly could be cumbersome and you have to repeat a lot of template code time and again, and handle the HTTP status code and response carefully. SDK provides the set of lanuage binds in Java, Python, Go and C++ etc, and it can help build the HTTP requests and make the API calls, allowing us to access OpenStack in a manner consistent with high-level language standards such as make function calls.

 

CLI, OpenStack also provied the command-line interfaces with OpenStackClients, along with the natvie Python API bindings for Compute,Identity,Image,Object Store and Block Storage etc. We have used CLI many times during the Openstack installation and configuration.

OpenStack API: Making the RESTful API calls to OpenStack

https://developer.openstack.org/api-guide/quick-start/

// To authenticate with OpenStack KeyStone

// and get the token and service catalog via cURL or Postman

URL: http://10.0.0.11:5000/v2.0/tokens

Method: POST

Headers: Content-Type:application/json

Data: {"auth":{"tenantName":"admin","passwordCredentials":{"username":"admin","password":"ipcc2014"}}}  

 

Response:

The project id: "448c4d583e1240ee93b0800c382fd494";

The token: "gAAAAABZ9hCPjL9D6uz3lqR6spcDgBWaR28VtkVZfGzcVTh-pPXPzvUElKuHLgiAekm7gVObUDGVjXdfM6ZqslYynFIf2aRKLdKZuP2W3Ps8yw70ZRISCFtRxN8wsOh-iEc1-ZUm0rqd9co_m7xSQTqRzFNtKKpAftAHbZYb58d0H2YofvHSUAk"

// To make HTTP requests to Nova with the received token and project ID, and get the instance info

URL: http://10.0.0.11:8774/v2.1/448c4d583e1240ee93b0800c382fd494/servers

Method: GET

Headers:

Accept:application/json

Content-Type:application/json

X-Auth-Token:gAAAAABZ9hCPjL9D6uz3lqR6spcDgBWaR28VtkVZfGzcVTh-pPXPzvUElKuHLgiAekm7gVObUDGVjXdfM6ZqslYynFIf2aRKLdKZuP2W3Ps8yw70ZRISCFtRxN8wsOh-iEc1-ZUm0rqd9co_m7xSQTqRzFNtKKpAftAHbZYb58d0H2YofvHSUAk

 

Response:

OpenStack SDK(Python): Making functional calls with the native Python bindings

https://wiki.openstack.org/wiki/SDKs

https://wiki.openstack.org/wiki/OpenStackClients

 

// To install pip and python-openstackclient

sudo apt-get installpython-pip

sudo apt-get install curl

sudo pip installpython-keystoneclient

sudo pip installpython-novaclient

sudo pip installpython-glanceclient

 

// To list all the images from OpenStack Glance

import keystoneclient.v2_0.client as ksclient

import glanceclient

 

# To authenticate with Keystone and get thetoken and servive catalog

keystone =ksclient.Client(auth_url="http://10.0.0.11:5000/v2.0",

                                              username="admin",password="ipcc2014",tenant_name="admin")

print(keystone.auth_token)                

 

# To get the service endpoint of Glance

endpoint = keystone.service_catalog.url_for(service_type=‘image‘,

                                            endpoint_type=‘publicURL‘)

# To list all the images

glance = glanceclient.Client(‘2‘,endpoint,token=keystone.auth_token)

for image in glance.images.list():

                print(image.id+ " " + image.name)

 

// To upload the image to OpenStack Glance

import keystoneclient.v2_0.client as ksclient

import glanceclient

 

def get_keystone_creds():

    d ={}

   d[‘username‘] = ‘admin‘

   d[‘password‘] = ‘ipcc2014‘

   d[‘auth_url‘] = ‘http://10.0.0.11:5000/v2.0‘

   d[‘tenant_name‘] = ‘admin‘

   return d

creds = get_keystone_creds()

keystone = ksclient.Client(**creds)

 

endpoint = keystone.service_catalog.url_for(service_type=‘image‘,endpoint_type=‘publicURL‘)

glance =glanceclient.Client(‘2‘,endpoint,token=keystone.auth_token)

 

newimage =glance.images.create(name="cirros_0_3_5", disk_format="qcow2",container_format="bare")

glance.images.upload(newimage.id,open(‘cirros-0.3.5-x86_64-disk.img‘, ‘rb‘))

 

// To list all instances

from novaclient import client

 

# Project ID: 448c4d583e1240ee93b0800c382fd494

nova = client.Client(2,‘admin‘,‘ipcc2014‘,‘448c4d583e1240ee93b0800c382fd494‘,‘http://10.0.0.11:5000/v2.0‘)

print(nova.servers.list())

print(nova.flavors.list())

  

OpenStack CLI: Manipulating OpenStack directly with commands

https://docs.openstack.org/python-openstackclient/latest/cli/

AWS SDK(Python): Making functional calls

 https://boto3.readthedocs.io/en/latest/guide/quickstart.html

 

// Install the latest Boto 3 release via pip

pip install boto3

 

// Set up authentication credentials and the defaultregion

// You need to create a AWS account

 

gedit~/.aws/config

gedit ~/.aws/credentials

// To list all the buckets and keys from AWS S3

import boto3

import botocore

 

s3 = boto3.resource(‘s3‘)

       

for bucket in s3.buckets.all():

   print(bucket.name)

 

for bucket in s3.buckets.all():

                                print(bucket.name)

                                forkey in bucket.objects.all():

                                                                print("----" + key.key)

 

# To create a bucket and upload the object

‘‘‘

s3.create_bucket(Bucket=‘xgf20171028‘,CreateBucketConfiguration={

   ‘LocationConstraint‘: ‘us-west-2‘})

data = open(‘test001.jpg‘, ‘rb‘)

s3.Bucket(‘xgftemp20170520‘).put_object(Key=‘test001.jpg‘,Body=data)

‘‘‘

Develop a simple Hybrid Cloud application

 

In the application, we will download the img file object from AWS S3, creat an image in OpenStack Glance and launch several instances with the image to simulate the data exchange between the Private and Public Cloud.

################################################################ 

# For AWS

import boto3

import botocore

 

# For OpenStack

import keystoneclient.v2_0.client as ksclient

import glanceclient

from novaclient import client

################################################################

s3 = boto3.resource(‘s3‘)

 

print("To list all the buckets and keys in AWS S3")

for bucket in s3.buckets.all():

                                print(bucket.name)

                                forkey in bucket.objects.all():

                                                                print("----" + key.key)

 

BUCKET_NAME = ‘xgf20171028‘

KEY = ‘cirros-0.3.5-x86_64-disk.img‘

 

print("To download the img object from AWS S3:")

try:

   s3.Bucket(BUCKET_NAME).download_file(KEY,‘cirros-0.3.5-x86_64-disk.img‘)

   print( "cirros-0.3.5-x86_64-disk.img is just downloaded from AWS S3to the local HD")

except botocore.exceptions.ClientError as e:

    ife.response[‘Error‘][‘Code‘] == "404":

       print("The object does not exist.")

   else:

       raise      

################################################################

keystone =ksclient.Client(auth_url="http://10.0.0.11:5000/v2.0",username="admin",password="ipcc2014",tenant_name="admin")

 

# get the endpoint of image service

endpoint =keystone.service_catalog.url_for(service_type=‘image‘,endpoint_type=‘publicURL‘)

print(‘the endpoint of OpenStack Glance: ‘ +endpoint) 

 

glance =glanceclient.Client(‘2‘,endpoint,token=keystone.auth_token)

 

print( "To list all the images inGlance:")

for image in glance.images.list():

                print(image.id+ " " + image.name)

 

# Tocreate a new image by glance with what was just downloaded from AWS S3 

newimage =glance.images.create(name="cirros_0_3_5", disk_format="qcow2",container_format="bare")

glance.images.upload(newimage.id,open(‘cirros-0.3.5-x86_64-disk.img‘, ‘rb‘))

print( "cirros-0.3.5-x86_64-disk.img isjust uploaded to OpenStack Glance")

   

print( "To list all the images inGlance:")   

for image in glance.images.list():

                print(image.id+ " " + image.name)

################################################################

 # the project ID of admin:‘448c4d583e1240ee93b0800c382fd494‘

nova =client.Client(2,‘admin‘,‘ipcc2014‘,‘448c4d583e1240ee93b0800c382fd494‘,‘http://10.0.0.11:5000/v2.0‘)

 

print( "To list all the VMs in OpenStack:")   

print(nova.servers.list())

 

# xgf_provider ID:995ccb97-fa06-403b-961c-5c11d8b24067

net_id = "995ccb97-fa06-403b-961c-5c11d8b24067"

nics = [{"net-id": net_id,"v4-fixed-ip": ‘‘}]

flavor =nova.flavors.find(name="m1.nano")

 

print( "To create 3 VMs:") 

nova.servers.create("xgf101",newimage, flavor, nics=nics)

nova.servers.create("xgf102",newimage, flavor, nics=nics)

nova.servers.create("xgf103",newimage, flavor, nics=nics)

 

print( "To list all the VMs inOpenStack:")   

print(nova.servers.list())

################################################################

                           

本文出自 “8483144” 博客，请务必保留此出处http://8493144.blog.51cto.com/8483144/1977277

Cloud实战:Manage and Manipulate OpenStack with CLI, SDK and API

标签：openstack cli sdk api aws hybrid cloud

原文地址：http://8493144.blog.51cto.com/8483144/1977277