掌握OpenStack设计模式：打造生产级云基础设施

需积分: 10 52 浏览量更新于2024-07-20 收藏 6.69MB PDF 举报

"《应用于实践的OpenStack设计模式》是一本2016年由Apress出版的专业书籍，由Uchit Vyas撰写，旨在帮助读者理解和掌握在实际生产环境中利用OpenStack组件构建可控制且定制化的云基础设施设计方法。本书深入浅出地介绍了OpenStack的基础知识，引导读者如何将应用程序流程与OpenStack组件相结合，探讨了组件和架构设计模式的设置过程。书中重点讲解了如何使用存储管理和计算来映射用户请求和分配，强调了高可用性和原生集群管理的重要性，提供了最佳实践指南。作者还详细讨论了OpenStack网络组件的解决方案，以减少延迟，优化组件间的通信，并促进与传统应用的无缝集成。这些设计策略对于现代云基础设施的设计者，包括经验丰富的解决方案架构师、DevOps专业人员以及系统工程师和分析师来说，是非常有价值的资源。读者将通过本书学习到： 1. 现代云基础设施设计的最新技术和技巧 2. 复杂应用程序基础设施设计的实战方案 3. 对OpenStack各种组件的深入理解，包括它们在支持现有或新云基础设施中的采纳和商业影响分析 4. 如何利用OpenStack特定组件与现有工具链集成，以实现敏捷开发和快速交付此外，本书还提供ISBN-13（纸质版）和电子版的识别码，以及版权和使用许可信息，确保了读者合法使用和传播内容。书中可能包含商标名称、标志和图像，这些都需要尊重知识产权。《应用于实践的OpenStack设计模式》是任何寻求在实际环境中高效利用OpenStack的专业人士不可或缺的参考指南。"

Chapter 1 ■ Designing Your First ClouD with openstaCk

Today, the most common issues persisting in the cloud world are to achieve agility, speed, and service

uptime up to 100%. All major datacenter solution provider companies are trying to get ready for the future’s

next-generation datacenter providers. Even in cloud, most of the public cloud vendors have expensive IT

systems; however, in the next couple of years, these systems will be out of date.

The major changes for the new-generation datacenter has evolved into the new model approach and

adoption for the provisioning and deployment of new software and hardware. To handle the new age of

cloud load, modern datacenters have to enable the multi-tenant model for scaling the approach. It is a huge

step in datacenter technology to shift their way of handling an entire infrastructure.

The next generation of Automation and DevOps for infrastructure has allowed system administrators

and operators to deploy and deliver a fully automated infrastructure within a fraction of the time. The

next-generation datacenter and automation will reduce all of the infrastructure components including

storage, network, compute, etc., as a single scalable and agile unit. So it’s the administrator’s responsibility

to code the infrastructure as per-environment requirements. To perform and adopt the next-generation

cloud mechanism where OpenStack comes as the ringmaster here to support next-generation datacenter

Operating System. The ubiquitous influence of OpenStack has been felt by many global giant cloud

enterprises like Rackspace, Red Hat, and Cisco, to name but a few. Nowadays most of the cloud services

providers are big giants running very large autoscaled private clouds based on OpenStack for their

customers and internal units.

Finally, based on what OpenStack can do, why should you use it? Well, OpenStack has very rich

community support, enhanced documentation, and elaborate tutorials. Let’s jump into the OpenStack

design part and its architecture.

Ultimately, the goal of this chapter is to get you from where you are today to the point where you can

confidently build a private cloud based on OpenStack with your own design choice.

This chapter will cover the following points:

• Getting familiar with the architecture of OpenStack and the way its different core

components co-relate with each other

• Learning how to design an OpenStack environment by choosing the right core

services for your environment

• Designing the first OpenStack architecture for a large-scale environment while

bearing in mind that OpenStack can be designed in numerous ways

At the conclusion of these first chapters, you will have a solid base of ways to plot your essential

functions and infrastructure modules by placing the specifics under the OpenStack lid. You will also

understand how OpenStack components work together and be prepared for the next step of our exploration

by entreating an OpenStack ecosystem in an intuitive foundation with best practices.

Though the attractive part of OpenStack might be to construct your own cloud, there are numerous

ways to achieve this purpose. Possibly the meekest of each is an appliance-style clarification. You grip an

appliance, unload it, plug in the power and the network, and watch it transmute into an OpenStack cloud

with nominal further configuration. Rarely, if any, other open source cloud commodities have such turnkey

possibilities. If a turnkey resolution is fascinating to you, take a glance at Nebula cloud.

Nevertheless, hardware selection is crucial for various applications, so if this pertains to you, know

that there are some software allocations presented that you can ride on servers, storage, and network

commodities of your choosing. Recognized giants named Red Hat and SUSE offer enterprise OpenStack

explanations and support.

Chapter 1 ■ Designing Your First ClouD with openstaCk

To object, deploy, and configure OpenStack, architects must recognize the rational design. Before

jumping into the architecture of OpenStack, we need to refresh the basic concepts and usage of each core

module. In order to get a clear understanding of how it plants, it will be applicable to first sketchily parse the

building chunks that mark its grind.

Supposing that you have previously mounted OpenStack or even positioned it in a small or medium-

sized ecosystem, let’s place the fundamental modules under the lid of OpenStack podium and go a little

more advanced by taking the use cases and asking the following: what is the tenacity of such a component?

1.1 Keystone

From an architectural viewpoint, Keystone offerings are the easiest type of service in the OpenStack

alignment. It is the primary module that stipulates identity tune-up, and it assimilates meanings for

authentication, catalog services, and policies to register and accomplish different tenants and users in the

OpenStack endeavors. The API requests concerning OpenStack services are being managed by Keystone to

confirm that the right user or service is able to exploit the entreated OpenStack service. Keystone achieves

plentiful validation apparatuses such as username/password as well as a token authentication-based

system. Moreover, it is possible to incorporate it with an accessible back-end directory such as Lightweight

Directory Access Protocol (LDAP) and the Pluggable Authentication Module (PAM).

A comparable real-life paradigm is a city resolute. You can grip a travel card and turnover by traveling

a definite number of places during a certain period of time. Before you start traveling, you have to test for

the card to gain entrance to the city at the main gates of the city domiciles. All the times that you would

like to attempt a new place for a visit, you must check in at the place gate machine. This will produce a

request, which is diagrammed to a central authentication system to pattern the validity of the card and its

permit, to profit the entreated place. By comparison, the token in Keystone can be competed to the travel

card excluding that it does not ebb anything from your request. The identity service is being measured as a

dominant and communal authentication system that feeds access to the consumers.

Horizon

Heat Templates

(Orchestration

)

Keystone

(Identity)

Neutron

(Networking)

Cinder

(Block Storage)

Nova

(Compute)

Glance

(Image)

Swift

(Object Storage)

Ceilometer

(Telemetry)

GUI

Stack Creation

Authentication

Persistent

Storage

Authentication

(OpenStack

Dashboard

or GUI)

Figure 1-1. OpenStack request workflow

Chapter 1 ■ Designing Your First ClouD with openstaCk

1.2 Swift

Swift is a highly available, distributed, eventually consistent object/blob store. Though it was succinctly

stated that Swift would be made obtainable to the operators beside the OpenStack modules, it is fascinating

to grasp how Swift has endowed what is signified to be cloud storage. Most of the businesses in the last

period did not fleece their doubts almost at precarious part of the IT groundwork — the storage where the

statistics are alleged. Thus, obtaining exclusive hardware to be in the innocuous zone had befitted a practice.

There are continually assured encounters that are handled by storage concocts and with no uncertainty,

one of these tests include the job of curtailing interruption while accumulating the data readiness.

Notwithstanding the upswing of many keen explanations for storage systems throughout the last limited

ages, we tranquilly demand to make deviations to the conventional fashion. Make it hazy! Swift was initiated

to accomplish this task.

We will consensus the givens distressing to the Swift strategy for subsequent, but you should endure in

communication that Swift is an object stowing software, which has a number of notable points:

• No dominant intelligence signposts and no Single Point Of Failure (SPOF)

• Healing designates auto-recovery in circumstance of disaster

• Particularly ascendable for hefty petabytes store and admittance by surpassing

matching

• Improved enactment, which is accomplished by dissemination of the freight over the

storage nodes

• Reasonable hardware can be utilized for superfluous storage clusters

1.3 Glance

The employment of Glance is to be an image registry. Starting here, we can accomplish how OpenStack has

secured the best approach to being further segmental and approximately coupled key component standards.

Utilizing Glance to heap virtual disk pictures is a choice. From a compositional level, joining added dynamic

approaches to a demand image report by means of the image storage Programming interface offered by

Look over a self-overseeing image storage back end, for example, Quick conveys further refreshing working

and taught framework hidden luxuries. With this strategy, a client will be savvy to record another virtual disk

image, on occasion, to stream it from an amazingly available and surplus store. On this adjusted stride, as a

specialized mechanic, you may give the appearance of another test execution.

1.4 Cinder

Truly, the administration of the persistent block storage can be incorporated into OpenStack by exercising

Cinder. Its key competence to extant block-level storage delivers raw volumes that can be assembled into

logical volumes in the filesystem and attached in the virtual machine (VM).

Certain landscapes that Cinder deals with are as follows:

• Volume supervision: this admits the making or removal of a volume

• Snapshot administration: this consents the design or erasure of a snapshot of

volumes

• You can attach or detach volumes from VMs

• You can replicate volumes

Chapter 1 ■ Designing Your First ClouD with openstaCk

• Volume construction from snapshots is achievable via Cinder

• You can duplicate images to volumes and vice versa

Numerous storage selections have been projected in the OpenStack fundamentals. Deprived

of hesitation, you may be questioning this: what kind of storage will be paramount for you? With a

policymaking technique, an evaluation of pros and cons must be completed. Table1-1 is a very naive table

that designates the transformation among the storage types in OpenStack to evade any misperception when

picking the storage administration decision for your forthcoming project:

Table 1-1. Storage Comparison

Measurement Storage Category

Object storage Block storage

Functioning - OK

Database storage - OK

Reinstating backup data OK OK

Setup for volume sources - OK

Persistence OK OK

Admission Wherever Inside VM

Image storage OK -

It is equally important to note that dissimilar Glance and Keystone facilities, Cinder landscapes are

conveyed by coordinating volume providers concluded the configurable operation driver’s designs such as

IBM, NetApp, Nexenta, and VMware.

If Cinder is proven to be an ideal solution or a replacement of the old nova-volume service that existed

before the Folsom release on an architectural level, it is important to know that Cinder has organized and

created a catalog of block-based storage devices with several differing characteristics. However, this is

obvious if we consider the limitation of commodity storage redundancy and autoscaling. Eventually, the

block storage service as the main core of OpenStack can be improved if a few gaps are filled, such as the

addition of these values:

• Quality of service

• Replication

• Tiering

The previously stated Cinder particular uncovers its non-seller lock-in trademark, where it is

conceivable to change the back end effortlessly or perform information relocation between two distinctive

storage back ends. In this manner, a superior storage outline design in a Cinder use case will bring an

outsider into the amusement of versatility.

1.5 Nova

As you may already know, Nova is the most unique center part of OpenStack. From an engineering level,

it is viewed as a standout among the most confusing parts of OpenStack.

Chapter 1 ■ Designing Your First ClouD with openstaCk

More or less, Nova runs continuously, which works together to react to a client demand into running

VM. How about we separate the blob image of Nova by accepting that its engineering as a disseminated

application needs organization so that assignments can be done between various segments?

1.5.1 nova-api

The nova-API segment acknowledges and reacts to the end client and figures the Programming interface

calls. The end clients or different parts speak with the OpenStack Nova Programming interface to make cases

by means of OpenStack Programming interface or EC2 Programming interface.

1.5.2 nova-compute

The nova-compute process part is basically a laborer daemon that makes and ends VM occurrences by

means of the hypervisor’s APIs (XenAPI for XenServer, Libvirt KVM, and the VMware Programming interface

for VMware).

1.5.3 nova-volume

The nova-volume part deals with the creation, joining, and separating of N volumes to process occurrences

(like Amazon’s EBS).

1.5.4 nova-network

The nova-network part acknowledges organizing undertakings from the Queue and after that plays out these

assignments to control the system, (for example, setting up connecting interfaces or changing the IP table

principles).

1.5.5 nova-scheduler

The nova-scheduler part takes a VM case’s solicitation from the line and figures out where it ought to run

(particularly which process server hosts it should keep running on). At an application engineering level, the

term planning or scheduler summons an orderly way to hunt down the best outfit for an offered framework

to enhance its execution.

Nova additionally gives console benefits that permit end clients to get to the console of the virtual case

through an intermediary: for example, nova-console, nova-novncproxy, and nova-consoleauth.

By zooming out the general parts of OpenStack, we find that Nova communicates with a few services:

for example, Cornerstone for confirmation, Look for images, and Horizon for the web interface. For instance,

the Glance connection is focal; the Programming interface procedure can transfer any question to Look,

while nova-compute will download images to dispatch images.

1.6 Queue

Queue gives a focal center point to pass messages between daemons. This is the place data is shared

between various Nova daemons by encouraging the correspondence between discrete procedures in a non-

concurrent way.

Any service can, without much of a stretch, speak with some other service by means of the API’s service.

One noteworthy position of the queuing framework is that it can cushion an extensive support workload.

As opposed to utilizing a RPC service, a queue framework can queue an extensive workload and give an

inevitable consistency.

剩余129页未读，继续阅读

vanridin

粉丝: 108
资源: 1187

掌握OpenStack设计模式：打造生产级云基础设施

英文原版-Applied OpenStack Design Patterns 1st Edition

OpenStack Trove(Apress,2015)

openstack atch design

OpenStack Days China 2016 联想高健：联想OpenStack云战略

OpenStack Days China 2016 OpenStack基金会Alan Clark：主题演讲

OpenStack Days China 2016 OpenStack基金会Jonathan Bryce：主题演讲

OpenStack Days China 2016 OpenStack基金会及英特尔苏义德：OpenStack 开放源码 开放思维

OpenStack Days China 2016 UnitedStack王为：VXLAN在OpenStack企业网实践

OpenStack Days China 2016 SUSE及OpenStack基金会Alan Clark:开幕致辞

OpenStack Days China 2016 九州云黄舒泉：OpenStack混合云之路

最新资源

OpenStack Days China 2016 OpenStack基金会及英特尔苏义德：OpenStack 开放源码开放思维