Continuous update. . . http://note.youdao.com/noteshare?id=cc7ac08e6e330360010d0a9460287267
This is migrated from Youdao Cloud Notes, and the format may be weird. . .
VMware
Citrix
Citrix System Architecture Design Basishttps:
//wenku.baidu.com/view/86e9196711661ed9ad51f01dc281e53a59025149.html
Citrix Desktop Cloud Service Solution
https://wenku.baidu.com/view/f65554273169a4517723a312.html?mark_pay_doc=2&mark_rec_page=1&mark_rec_position=2&clear_uda_param=1
AWS
AWS - Amazon's cloud computing service platform
Azure
Microsoft Azure is an open and flexible enterprise-grade cloud computing platform. Accelerate your development, increase productivity and save operational costs with IaaS + PaaS.
Openstack/CloudStack/OpenNebula/Eucalyptus
Common IaaS open source platforms include OpenStack, CloudStack, Eucalyptus, and OpenNebula. The earlier public status reports and community analysis found on the Internet show that OpenStack and CloudStack have far surpassed the development of OpenStack and CloudStack in the enterprise and community. For a comparison of the two, refer to the following articles:
About openstack, cloudstack, Eucalyptus, OpenNebula comparative analysis
http://blog.csdn.net/wood893278466/article/details/52945076
Openstack
OpenStack adopts a distributed architecture. The entire platform is divided into multiple module projects according to different functions. The projects communicate with each other through message queue middleware and RESTful API (where the certification project provides certification and service registration for other projects). Therefore, each project Each project can be independently deployed on different hosts, and the architecture is very flexible. OpenStack provides rich and powerful APIs, so that resources can be easily used and scheduled by users, and it also provides Amazon AWS (Amazon Web Services) compatible. API.
The basic concept and architecture diagram of OpenStackhttp ://blog.csdn.net/zjluobing/article/details/51489325
OpenStack provides an Infrastructure-as-a-Service (IaaS) solution through a set of related services. Each service provides an application programming interface (API) that facilitates this integration. Depending on your needs, you can install some or all of the services. The following table describes the OpenStack services that make up the OpenStack architecture:
|
Service
|
Code Name
|
Description
|
|
Identity Service
|
Keystone
|
User Management authentication, user management
|
|
Compute Service
|
Nova
|
Virtual Machine Management Virtual Machine Management, Scalable Compute Configuration Engine?
|
|
Image Service
|
Glance
|
Manages Virtual image like kernel image or disk image 镜像服务
|
|
Dashboard
|
Horizon
|
Provides GUI console via Web browser
|
|
Object Storage
|
Swift
|
Provides Cloud Storage Object Storage
|
|
Block Storage
|
Cinder
|
Storage Management for Virtual Machine Block Storage
|
|
Network Service
|
Neutron
|
Virtual Networking Management
|
|
Orchestration Service
|
Heat
|
Provides Orchestration function for Virtual Machine
|
|
Metering Service
|
Ceilometer
|
Provides the function of Usage measurement for accounting
|
|
Database Service
|
Trove
|
Database resource Management
|
|
Data Processing Service
|
Sahara
|
Provides Data Processing function
|
|
Bare Metal Provisioning
|
Ironic
|
Provides Bare Metal Provisioning function
|
|
Messaging Service
|
Zaqar
|
Provides Messaging Service function
|
|
Shared File System
|
Manila
|
Provides File Sharing Service
|
|
DNS Service
|
Designate
|
Provides DNS Server Service
|
|
Key Manager Service
|
Barbican
|
Provides Key Management Service
|
Complete deployment of CentOS7.2+OpenStack+kvm cloud platform environment (1)--basic environment construction https://www.cnblogs.com/kevingrace/p/5707003.html
Detailed graphic tutorial on installing and deploying private cloud based on openstack http://www.jb51.net/article/104511.htm
Additional Reading: Virtualization Technology
Introduction to Xen, OpenVZ, KVM, Hyper-V, VMWare virtualization technology
http://blog.csdn.net/myweishanli/article/details/24184489
http://blog.csdn.net/myweishanli/article/details/24184489
Additional reading: Storage
The essential difference between block storage, file storage, object storage and distributed file storage system
http://blog.csdn.net/enweitech/article/details/51445087
Block storage and file storage are two mainstream storage types that we are familiar with, while Object-based Storage is a new network storage architecture (which can be understood as a new storage type), based on object The storage technology device is the object storage device (Object-based Storage Device) referred to as OSD.
Generally speaking, all disk arrays are block-based, and all NAS products are file-level storage.
1. Block level concept:
Block level refers to the sector-based, one or I consecutive sectors to form a block, also called a physical block. It is between the file system and block devices (eg: disk drives).
2. File-level concepts:
The file level refers to the file system. A single file may be composed of one or more logical blocks, and the logical blocks are distributed discontinuously. The logical block is greater than or equal to an integer multiple of the physical block,
3. Diagram of the relationship between physical blocks and file systems:
Mapping relationship: sector→physical block→logical block→file system
File-level backup:
File-level backup means that when specifying certain files for backup, the logical blocks of each file will be searched first, followed by physical blocks, because logical blocks are scattered on physical blocks, and physical blocks are also scattered on different sectors. You need to search down layer by layer, and finally complete the entire file copy. File-level backup is time-consuming, inefficient, ineffective in real-time, and takes a long time to backup. During incremental backup, a small part of a single file is modified, not only the modified part, but the entire file.
Block-level backup:
Block-level backup refers to physical block replication, which has high efficiency, strong real-time performance, and short backup time. During incremental backup, only modified physical blocks are backed up.
Distributed storage is not a storage type, but means that the entire storage system is composed of multiple storage servers, which are connected through a network (optical fiber, Ethernet, IB network, etc.), and are managed by a unified management center. resource. In terms of specific storage methods, you can use block storage (DAS, SAN) or file storage (NAS), but the method that combines the dual advantages of SAN and NAS is object storage.
Further reading:
|
[Block storage] Typical equipment: disk array, hard disk block storage mainly maps the entire raw disk space to the host for use, that is, for example, there are 5 hard disks in the disk array (for the convenience of description, assuming that each hard disk is 1G), then you can N logical hard disks are logically divided by various methods such as partitioning logical disks, doing Raid, or LVM (logical volumes). (Assume that the divided logical disks are also 5, each of which is 1G, but these 5 1G logical disks have completely different meanings from the original 5 physical hard disks. For example, in the first logical disk A, the first logical disk may be the first one. 200M is from physical hard disk 1, and the second 200M is from physical hard disk 2, so logical hard disk A is a hard disk logically fictionalized by multiple physical hard disks.) Then block storage will use mapping to map these logical disks to The host, the operating system on the host will recognize that there are 5 hard disks, but the operating system cannot distinguish whether it is logical or physical. There is no difference to the operating system, at least there is no difference in the perception of the operating system. In this way, the operating system also needs to partition and format the mounted bare hard disk before it can be used, which is completely the same as the way of the built-in hard disk of the ordinary host. Advantages: 1. The advantage of this method is, of course, that data is protected by means of Raid and LVM. 2. In addition, multiple cheap hard disks can be combined to form a large-capacity logical disk to provide external services and increase the capacity. 3. When writing data, since it is a logical disk composed of multiple disks, several disks can be written in parallel, which improves the read and write efficiency. 4. In many cases, the block storage adopts the SAN architecture network, the transmission rate and the encapsulation protocol, so that the transmission speed and the read and write rate are improved. Disadvantages: 1. When using SAN architecture for networking, it is necessary to purchase additional Fibre Channel cards for the host, as well as fibre switches, which are costly. 2. The data between hosts cannot be shared. If the server is not clustered, the block storage bare disk is mapped to the host. After formatting and using, it is equivalent to the local disk for the host, so the local disk of host A cannot be used at all. It is used by host B and cannot share data. 3. It is not conducive to data sharing between hosts of different operating systems: Another reason is that because the operating systems use different file systems, after formatting, the data between different file systems cannot be shared. For example, one installed WIN7/XP, the file system is FAT32/NTFS, and Linux is EXT4, EXT4 cannot recognize the NTFS file system. Just like an NTFS-formatted USB flash drive, inserted into a Linux notebook, it cannot be recognized at all. So not conducive to file sharing. [File storage] Typical equipment: FTP, NFS server In order to overcome the above problem that files cannot be shared, file storage is provided. There are also software and hardware integrated devices for file storage, but in fact, if you take a server/laptop, as long as you install the appropriate operating system and software, you can set up FTP and NFS services. After such services are installed, the server is the file. kind of storage. Host A can directly upload and download files to the file storage. Unlike block storage, host A does not need to format the file storage, because the file management function is already handled by the file storage itself. Advantages: 1. Low cost: just one machine is enough, and ordinary Ethernet is enough, no dedicated SAN network is needed at all, so the cost is low. 2. Convenient file sharing: For example, host A (WIN7, NTFS file system), host B (Linux, EXT4 file system), if you want to copy a movie to each other, it is not possible. Add a host C (NFS server), then you can first copy A to C, and then copy C to B and it will be OK. (The example is superficial, please forgive me...) Disadvantages: Low read and write rate, slow transfer rate: Ethernet, slow upload and download speed, in addition, all read and write needs to be undertaken by a hard disk in a server, compared to disk arrays Dozens or hundreds of hard disks can be read and written at the same time, and the speed is much slower. [Object storage] Typical equipment: distributed server with built-in large-capacity hard disk The most commonly used solution for object storage is multiple services The device has a built-in large-capacity hard disk, and then installs the object storage software, and then creates a few additional services as management nodes, and installs the object storage management software. The management node can manage other servers to provide read and write access to the outside world. The reason for the emergence of object storage is to overcome the shortcomings of block storage and file storage, and to promote their respective advantages. In short, block storage is fast to read and write, which is not conducive to sharing, and file storage is slow to read and write, which is conducive to sharing. Can you get a fast read and write? It's about sharing. Hence the object storage. First, a file contains attributes (termed metadata, metadata, such as the file size, modification time, storage path, etc.) and content (hereinafter referred to as data). In the past, a file system like FAT32 directly stored the data of a file together with metadata. The storage process first broke the file according to the minimum block size of the file system (such as a 4M file, assuming that the file system requires a block of 4K). , then break the file into 1000 small pieces), and then write it into the hard disk, without distinguishing data/metadata in the process. And each block will finally tell you the address of the next block to be read, and then continue to follow the map in this order, and finally complete the reading of all blocks in the entire file. In this case, the read and write rate is very slow, because even if you have 100 robotic arms reading and writing, you can only know where the next block is because you can only read the first block, which is equivalent to only one. The robotic arm is actually working. The object storage separates the metadata, and the control node is called the metadata server (server + object storage management software), which is mainly responsible for storing the attributes of the object (mainly, the data of the object is scattered and stored in the distributed servers). The information in the file), while other distributed servers responsible for storing data are called OSDs, which are mainly responsible for storing the data part of the file. When a user accesses an object, it will first access the metadata server. The metadata server is only responsible for reporting which OSDs the object is stored in. Assuming that the feedback file A is stored in the three OSDs B, C, and D, the user will directly access the three OSD servers again. to read data. At this time, because 3 OSDs transmit data at the same time, the transmission speed is accelerated. The greater the number of OSD servers, the greater the improvement in read and write speed. In this way, the purpose of fast read and write is achieved. On the other hand, the object storage software has a special file system, so the OSD is equivalent to a file server to the outside world, so there is no difficulty in file sharing, and the problem of file sharing is also solved. Therefore, the emergence of object storage combines the advantages of block storage and file storage well. Finally, why does object storage have the benefits of both block storage and file storage, yet use block storage or file storage? 1. There is a class of applications that need to store direct bare disk mapping, such as databases. Because the database needs to store the bare disk and map it to itself, and then format the bare disk according to its own database file system, it is not possible to use other files that have been formatted as a certain file. system storage. Such applications are better suited to use block storage. 2. The cost of object storage is still higher than that of ordinary file storage, and special object storage software and large-capacity hard disks need to be purchased. If the data volume requirements are not massive, but only for file sharing, it is better to use the form of file storage directly, which is cost-effective.
|