A, IP network infrastructure
(A) internet Introduction to History
1960 (ARPANET) -1970 (Internet concept was born, TCP / IP birth) -1980 (CSNET NSFNET) -1990 (ANASNET) - now (Internet)
(B) What is the Internet?
1, the network: the set of communication devices 2 connected to communicate with.
2, Internet: 2 or a plurality of devices communicate with each other using TCP / IP protocol interconnect.
Between the device and the device constituting the network, through the Internet network and the network of TCP / IP protocol are interconnected through the network interconnection media.
(C) data communication protocol basis:
1, OSI RM: Development Systems Interconnection Reference Model (open
system interconnection reference model) - seven layer protocol architecture - physical layer, a data link layer, network layer, transport layer, session layer, presentation layer, application layer
2, TCP / IP protocol suite - five-layer protocol architecture - physical layer, a data link layer, network layer, transport layer, application layer
(D) TCP / IP layers function
1-3: Main layer data communications
1, the physical layer: 1, the voltage defined interfaces, standard cable, the transmission distance. 2, the transmission bit stream. 3, defines the transmission medium: coaxial cable, twisted pair, fiber optic, radio waves.
Physical layer mainly equipment: T4 HUB hub (exchange), the relay repeater
2, link layer number: 1, the physical MAC address defined. 2, the definition of the link parameters of the flow control parameters. 3, error verification.
Major equipment: Layer 2 / Ethernet switches, bridges.
Physical address (MAC) Introduction
Globally unique physical address for 48 binary (dotted hexadecimal), the first 24 numbers for the manufacturers, after 24 vendors are automatically assigned.
Physical address in the physical memory package (card), after running out of memory to run the call.
3, the network layer:
Addressing equipment for three main functions:
(1) defines a network layer address. (2) addressing and forwarding. Logical addressing ---
Equipment: routers, three switches. --- According to IP address addressing.
Network Protocol: IP protocol (ip network) IPX protocol (Novell network).
ip address
32 binary composition (decimal point).
In 2 parts: the network address and host address. (Hierarchy)
Functions: public address, private address:
10.0.0.0 -10.255.255.255
172.16.0.0 -172.31.255.255
192.168.0.0-192.168.255.255
NAT translation: private network addresses across the public network address translation.
Background: The global IPv4 resources have been allocated in February 2011, to meet the demand for more Internet users need to convert the private address for the public network address for a visit.
Ways: conversion function for the UDP, TCP protocol packet header address and port number. (Refer to information on the Internet protocol packets)
Between private and public networks: deployment location.
4, the transport layer:
(1) communications between processes and processes. (2) flow control and error mechanisms.
Error mechanisms: the link layer is mainly connected to each other directly control the flow point of the two devices (partial flow control). Transport layer mainly flow control between a process and the process.
Transport Layer: UDP, TCP
UDP protocol:
Connectionless protocol, also known as transparent protocol, also located in the transport layer.
TCP protocol:
Located in the transport layer, provides reliable byte stream service. Called byte stream service (Byte Stream Service) refers, for convenience of transmission, the data is divided into chunks of data packets segment (segment) of the operator or manager. And may
By means of the transport service, capable of accurate and reliable data to pass to the other side. That is a large TCP protocol data only then split for easier data transfer, data and TCP protocol can confirm whether the final delivery to the other party. Therefore, TCP connection is equivalent to
Two tubes (one for the server to the client for a client to server), the data transmission pipe which is a bytecode transmission, the transmission is ordered, each byte is one of a transmission.
Difference between the two:
1) TCP provides a connection-oriented transport, must first establish a connection (three-way handshake) before communication; the UDP provides a connectionless transport, before the communication without establishing a connection.
2) TCP provides reliable transmission (ordered, error-free, is not lost, will not be repeated); the UDP provides unreliable transport.
3) TCP byte stream oriented transport, so that it can be divided into groups of information, and the receiving end of its restructuring; is the UDP datagram-oriented transmission, no packet overhead.
4) TCP provides flow control and congestion control mechanisms; the UDP does not provide congestion control and flow control mechanisms.
Original link: https: //blog.csdn.net/striveb/article/details/84063712
Two, IP routing techniques are described
IP Routing: between Client Access servers through network devices.
1, the routing table: display ip route ip route table view
Destination address, the next hop address, protocol type, by the preference value (priority), the value Cost (metric), the interfaces
10.3.1.0 - (E1 E0) -10.1.2.0- (E0 E1) -10.2.1.0- (E0 E1) - 10.4.1.0
2, routing protocol
(1) direct direct routes
(2) non-direct routes: static routing static, dynamic routing
1) interior gateway protocol: the role of the autonomous system AS internal
rip、ISIS、ospf
2) External Gateway Protocol: acting between autonomous systems AS
BGP
3, routing algorithm:
(1) distance vector algorithm: EGP RIP
Routers send routing information to obtain the minimum number of hops routes to neighbors. Neighbors can only get routing information, can not form a network topology.
(2) Link Protocol Algorithm: ISIS OSPF
Acquiring the destination host until the entire route, network topologies FIG dynamic routing table count LSP neighbor route to the router occurs, the shortest path through the SPF number generating algorithm.
Third, the definition and history of the development server
Server: a PC running relatively high speed, high load, the higher the price.
Server Features: the CPU high speed, reliable operation for a long time, a powerful I / O (input / output) data throughput.
Server Infrastructure Services business: instant messaging, web services, virtualization, document printing.
Server service scenario: the Internet, Big Data, HPC high-performance science and technology scene.
Industry 4.0
IT Information Technology dragging the 1940s to the third technological revolution. On the basis of the third technological revolution, information technology, big data and Internet of things combined intelligent plant, intelligent production formed a key industry 4.0.
Server is a direct role in promoting industrial 3.0 4.0.
In 1946 the first day of computers, first mainframe 1964, 1965 and the first minicomputer, the 1990s UNIX Server 1989 inter X86 server,
Fourth, the server type
Shape classification: Tower, rack, blade, high density
Performance: The number of CPU. Single, double, four or more multiple servers.
Instruction Set Category: RISC RISC (non-X86), complex instruction set CISC (X86)
Categories: databases, applications, WEB, access, file
Components: CPU, memory, hard drives
V. server technology and architecture
Three major trends: sacle-up vertical expansion, hyper-converged over converged infrastructure, sacle-out scale
1, sacle-up: enhance the performance of a single service.
Features: high reliability, high availability, high scalability.
Applications: high performance transaction types, HPC-critical applications and high-performance computing.
Areas: telecommunications billing, scientific research, weather analysis, etc.
2, sacle-out: coordination of multiple servers is completed, the high number of concurrent improve performance.
Features: low cost, high density, and energy saving.
Application: Large data centers.
Areas: big data analysis, public cloud, web applications and other business clusters.
3, hyper-converged: storage, computing, networking, unified management integration.
Features: Easy to use, performance optimization, high speed overall integration.
Applications: high performance data analysis, database integration, cloud computing resource platform, integrated data center.
Workstations: get up early application is a standalone application to deploy all applications on a single server, or provide computing services to a business or a few users.
4, C / s Application Architecture
Architecture: a shared database for multiple applications client application deployment architecture composed of two layers.
Advantages: a shared database, a plurality of terminal docking application server, service data to achieve a cooperative operation, shared business and uniform preservation.
Drawbacks: application deployment and maintenance of complex configuration, software upgrades, PC client upgrade and maintenance needs one by one, is not conducive to flexible application deployment, is not conducive to large-scale client application and promotion.
5, B / s application architecture.
Architecture: three-tier architecture, application server software, install the server in the background.
Pros: Users can obtain the corresponding service connection to log server through a Web browser, connect different applications in different application server through a Web browser. Simple to use, maintenance-free, industry
Business application software, system data can be centralized deployment and maintenance for large-scale application system deployment and application.
Networking architecture: B / s
Six top software server architecture
1, system installation services and service deployment
(1) Unix server
Unix operating systems: AIX Solaris HP-un11
(2) X86 server
X86 server operating systems: Linux, Windosw
(3) virtualization system
VMware FusionSpherr KVM server can be a physical simulation of multiple smaller virtual servers, it can provide a better resource efficiency through server virtualization, deployment and simplify management.
(4) main server database
Oracle, IBM.DB2, Mysql, Sql server, Mysybase, NPC gold positions, up to a dream database
(5) middleware (Java-based)
Provides operating and development environment for the upper layer application software, pre-reusable business functions module API interface
Business: Weblogic, WebSphere, Tuxedo, East Pass.
Open Source: Tomcat, JBoss
(6) The main server business applications
peoplesoft、ERP、CRM、HR、Email、logic APP、iFLEX、SIEBEL
, Custom development based on customer needs based on the database, middleware infrastructure platform.
(7) the correspondence between cloud services architecture
Saas: Business Application
Pass: databases, middleware
LaaS: servers, storage, networking and other hardware infrastructure, operating systems, virtualization
Seven, storage
1. The data stored in
Protection of important data stored carrier
2. The importance of data
Data provide support for business operations, analysis and decision-making.
3. Features
Structured data:
Unstructured data:
4. narrow storage
Floppy disks, CD DVD, hard disk, tape
The generalized memory
(1) a data storage device comprising a storage center hardware, system software, network storage, storage solutions.
(2) accessing the network server data storage hardware, software system provides management of data, the plurality of storage hardware and software combine to form a solution.
(3) meet high professional data management requirements, data integration, disaster recovery solutions
6. Business and storage
A sophisticated system for storing and managing business critical information.
Example: the daily consumption is saved to storage.
Important data to protect against data security software and storage. (Backup)
Eight, storage development
Development course and two drivers
- user requirements - Tech Tips
2, early storage
Integrated storage server configuration is low (CPU, memory, disk), hard disk performance, capacity and reliability can not meet the business development needs.
Solution: The hard disk separate from the server out of the formation of professional storage systems.
3, DAS storage form
JBOD: to enhance the capacity external hard disk, the database server is responsible for the management still exists overhead.
Intelligent: the integrated data management, access control and function, the server only need to focus on reading and writing data.
Application: VAS system operators.
4, SAN and NAS storage
Requirements: data sharing multiple services, unified data management and greater data service capacity and performance requirements.
（1）SAN
A first data server stores the file system format for an application performing data read and write.
Scenario: data storage structure of the database.
High-end storage: IBM DS8000 VSP Huawei 18,000
Terminal stores: for stable, high performance requirements of the scene. Used in corporate, government core class system.
(2) IN
Early user-oriented office data sharing.
Features: put the server file system for storage, the server over the network can share data.
Scenario: data storing unstructured data.
NAS gateways: hanging in front SAN storage. Applied to file sharing, VDI scenarios.
Large data mass storage: provides strong expansion of capacity (on behalf of EMC islion and Huawei oceanstor 9000). Apply to media assets library, video surveillance, high-performance computing.
(3) SAN / NAS storage applications
Image Information: Image Index information is used to query belong to structured data. Specific image files belonging to unstructured data.
(4) Unified Storage
SAN NAS binding to, an object to provide two or more functions in the same set of hardware and software basis.
Typical storage: FAS, OCEANSTOR V3,3Par belonging BEST MODE no additional NAS gateways. VNX need additional NAS gateways.
Nine, database basics
1, data concepts
Mounting a data storage management organizational structure of the data warehouse, the relational database into two categories, non-relational database
(1) relational database
Created on the relational model database is stored on a table-type line.
Features: strong consistency requirements, site, consistency, isolation, and of magnetic button.
Typical: oracle, DB2, sql server
Management system: a set of data organized according to a certain format to store and process data, mainly through the database management system implementation, all data access requests require a database DBMS processing.
DBMS features:
1, data definition. We must support a variety of data definition language processing and compiling, analyzing and corresponding data processing requirements.
2, data manipulation. Data in the database can be add, delete, change,.
3, optimize execution. Data Manipulation Language request must be optimized, devices, process. After optimization request operation performed under the control of the manager.
4, data security and integrity. Must monitor the user's request, it has refused to request the security integrity of corrupt data.
Relational Data Model
Table: basic storage object database. A collection of two-dimensional array, used to represent the relationship between the stored data.
Column: Description of the properties of the specific track. Having a data type and length.
OK: Ganso, the operation of the database table data is achieved by sql language.
Constraints: notno constraints, unique constraints
(2) non-relational databases
Data storage mode and does not require non-fixed form association, have their own characteristics with the API interface and relational data base needs of different available weak consistency, soft consistency and a final state.
Typical: Nosql
With the advent of web2.0 as well as business development, relational database difficult to deal with large scale and high concurrent SaaS business requirements.
Scene: simple data model, IT systems need more flexibility, higher performance requirements for the database, does not require a high degree of data consistency, for a given environment KI relatively easy to map complex values.
classification:
1, key-value store database. Will be used mainly to the hash table, the table has a pointer to a particular key and specific data.
Scene: content caching, access to high-load large amounts of data, but also for the logging system.
Pros: fast search speed.
Disadvantages: unstructured data, and can only be treated as a string or binary data.
2, column-store database. To deal with distributed storage of huge amounts of data, the key still exists, and there is a point to multiple columns.
Scene: a distributed file system.
Pros: Find fast, strong scalability, easily distributed expansion.
Disadvantages: relatively limited functions.
3, document database. Store semi-structured document format features.
Scene: web application.
Advantages: the data format is not strictly required, variable table structure, no need to define the table structure.
Disadvantages: query performance is not high, lack of a unified query syntax.
4, graph database. Flexible graphical model can be extended to multiple servers.
Scene: social networking, recommendation systems, focus on building relationships map.
Advantages: correlation algorithm using the structure of FIG. For example the shortest path addressing, N and the like of the lookup relation.
Cons: Very often it needs to be done to get the information needed to calculate the entire map, is not very good to do a distributed cluster program.