abstract
In order to protect, optimize and develop business, any business continuity plan must include reliable network infrastructure to meet the data protection needs of various institutions, enable applications to resume normal operation quickly, and ensure the continuity of user access. Cisco system? Based on the scalable and adaptive enterprise data center architecture, the company provides business continuity network solutions for enterprise data center managers. In addition to various user access technologies, it can also provide cost-effective communication between multiple data centers and recovery sites.
challenge
Enterprise data center supports many applications that are vital to business operation, and it is the focus of business continuity strategy of various institutions. If applications and data are unavailable, or users cannot access them, enterprises will not only lose revenue, but also reduce production efficiency. To make matters worse, this situation will also damage the impression of customers and partners on the enterprise, or be subject to legal proceedings. In order to achieve the management goal of protecting, optimizing and developing business, many IT organizations are rethinking their business continuity plans. Business continuity is the most concerned issue for enterprises that are highly sensitive to risks and threats, new regulations and the integration trend of IT and business resources in data centers.
At present, the infrastructure of many data centers is separated, which is not only difficult to manage and maintain, but also difficult to adapt to new technologies and applications. Moreover, in the case of interruption or disaster, it is difficult to implement consistent and appropriate protection for the isolated application environment, and the maintenance cost is very high.
In order to build a network that can support the business sustainable development strategy, data center and business continuity managers face many problems:
Tape backup has many limitations-as a traditional mainstream disaster recovery method, tape backup is slow, and it is difficult to recover key applications and data on time and meet the recovery time requirements. However, its role in maintaining integrity cannot be ignored.
High cost-Although the cost of business continuity technology has been declining in recent years, the cost of providing mirror software, redundant facilities and resources, telecommunications services and corresponding manpower is still high. Therefore, IT managers also need to use a series of solutions to control capital costs and operating costs, and ensure that enterprises can achieve business continuity goals.
Long latency-Because synchronous mirroring with zero data loss requires data to be written to local and remote locations before the transaction is acknowledged, it will affect the performance of the application. Transmission technologies such as MAN can effectively shorten the delay. Generally speaking, due to the limitation of light speed, for most real-time transaction processing applications, the actual distance between synchronous mirror sites can reach 200km.
High cost of network facilities-Although the bandwidth cost continues to decline, the mirroring and replication of large data centers require extremely high capacity. If the traditional dedicated line service is adopted, the cost will be very considerable.
User access-Business continuity plans usually focus on the availability of data and applications, but often ignore the need for users to access recovered data and applications.
Basic elements of business continuity
The business impact assessment points out the risks and potential costs associated with application downtime, data loss and user access interruption. Although the ideal situation is to prevent operation interruption and data loss to the greatest extent, it is either too expensive or unrealistic to achieve this goal. In fact, many applications allow some interruption, and only some critical business applications need to be available at any time. Risk assessment uses two business continuity metrics to evaluate applications:
Recovery Point Objective (RPO)- Determine the risk of data loss that the organization is willing to bear, ranging from zero to several minutes or hours;
Recovery Time Objective (RTO)-Quantify the maximum downtime of an application.
Although RPO and RTO indicators are very useful for data and application-level evaluation, Cisco Systems has proposed a third indicator, which is directly applicable to the Network Recovery Access Goal (RAO). RAO determines the time to restore users to the application, no matter from which point in time. Without RAO, the practical value of implementing RPO and RTO will be very limited. For example, during normal operation, branch users can access key business applications located in the main data center through the intranet WAN. If the application fails, users need to use fallback methods, such as quickly establishing a site-to-site virtual private network (VPN) connection through the Internet Service Provider (ISP) network, and establishing a connection with the standby data center that supports application recovery.
Organizations must consider the proximity of all backup data centers or sites. Because only by setting the backup data center or backup site far enough can we reduce the impact of disasters such as power failure, volcanic eruption, hurricane or earthquake.
Figure 1 Schematic diagram of business continuity solution
Using RPO, RTO and RAO indicators and distance requirements, business continuity managers can equip each application with technical solutions corresponding to their business roles (Figure 1). These solutions may contain one or more of the following options:
A highly available storage network-storage local area network (SAN) or network attached storage (NAS) solution can deal with local server failures because it can provide access to standby or clustered server systems to ensure continuous operation. * * * Enjoy the storage environment and use advanced storage system to ensure that disk failure will not affect the normal operation of the application.
Centralized backup-Regular backups are part of the standard operating procedures for most business applications. As the cheapest solution, centralized tape backup takes the longest recovery time, but it is very suitable for preventing data corruption. At present, because the cost of disk storage media is getting closer to that of tape storage, the disk-based "near-line" backup system is increasingly welcomed by enterprises because of its short backup time and recovery time.
Remote asynchronous replication-Replicating data changes to a remote system in another location. According to the importance of data, data changes can be copied regularly at the specified time every day, or asynchronously using storage systems, NAS devices, host systems or advanced software embedded in the network. The asynchronous nature of these solutions has strong advantages. Because distance does not affect performance, there is no delay limit. The recovery time of data replication is shorter than that of tape backup, but the asynchrony cannot guarantee the recovery of all data.
Synchronous disk copy and mirroring-For applications that require the fastest recovery and do not allow data loss, advanced software on storage systems, hosts or centralized storage networks can be used to implement synchronous disk copy and mirroring solutions. The transaction is not complete until all disk writes are synchronously replicated to the remote site through a high-performance network, thus eliminating any transaction loss. The limitation of this method lies in the amount of delay, which limits the actual distance between sites. This is especially important for write-intensive applications and online transaction processing applications such as enterprise resource planning and customer relationship management.
Data center mirroring-In order to achieve the highest level of continuity, more and more enterprises have established two main data centers and built an extended server cluster between the two sites. Some enterprises usually only use one main data center, and the other data center is only used in case of failure or disaster. But in fact, two data centers can be used as "hot" data centers or currently active data centers at the same time, that is, all transactions in one data center are mirrored to the other data center at the same time. The solution * * * contains two modes of operation. In the first operation mode, two data centers * * * share the transaction load of the application and are completely synchronized, and user traffic is distributed to two sites by using global load balancing. Mirroring the data center and the main data center can ensure that the failure recovery of any data center is transparent to users. In addition, this model can make better use of expensive investment and improve the return on investment. In the second mode, each data center will act as the main data center for some applications, while providing mirroring for applications in another data center. If one data center fails, the other data center will take over all its applications. Either way, data center mirroring can provide the highest level of business continuity.
Continuous user access and connectivity-Business continuity must include a RAO policy to keep users connected to applications when conditions change. Depending on the severity of the disaster, some access methods are transparent to users, while others need to be manually reconfigured. Technologies that are transparent to users include global load balancing between mirrored data centers, automatic path switching in case of link failure, and automatic recovery to standby components or devices. Technologies that need to be reconfigured include user connection modification. For example, if the local area network is damaged by a fire in the company building, users can use VPN connection for remote communication and access applications.
Cisco enterprise data center architecture
Cisco? Enterprise data center architecture is a harmonious network architecture, which can not only meet the requirements of convergence, business continuity and security of current data centers, but also meet the requirements of future data centers for service-oriented technologies and application computing technologies such as blade servers, virtualization, Web services and grids. Through this architecture, Cisco Systems, the global leader in data center network field, can provide appropriate infrastructure for IT and network managers and realize a complete and extensive business continuity strategy. Cisco enterprise data center is based on intelligent network, which can not only help enterprises to eliminate the risk of application interruption, but also help enterprises to make corresponding plans for scattered data centers and realize more advanced business continuity functions. By using tested and validated architecture, practical design best practices and configuration templates, Cisco can help IT managers adopt this architecture, reduce risks, shorten backup and recovery time and reduce investment requirements, which are universal and take into account the characteristics of each partner. With its flexibility, enterprises can deploy computing, storage and software technologies, which not only help to achieve enterprise goals, but also help to effectively implement new services and applications. After implementing this adaptive data center network architecture, IT organizations can better achieve their corporate goals of protecting, optimizing and developing their business. It can not only protect key applications and confidential data, but also improve the operation efficiency of data center, quickly establish a safe new application environment and effectively support new business processes. By using the highly flexible and effective adaptive data center network, enterprises can alleviate the competitive pressure, expand the market scale and speed up the listing process of new businesses, thus re-coordinating various resources and promoting the development of enterprises.
Cisco enterprise data center architecture includes three levels (Figure 2):
Infrastructure includes intelligent IP network infrastructure, intelligent storage network and data center interconnection;
Network system intelligence includes security, supply optimization, manageability and availability;
Embedded applications and storage services include storage virtualization, data replication and distribution, and advanced application services.
Figure 2 Business Continuity Network: Composition of Enterprise Data Center Architecture
Cisco Business Continuity Network Solution
Business continuity network is the key foundation of any business continuity strategy. It can not only provide the connection between the application and the storage/backup source, the connection between the main data center and the standby data center, but also optimize the user's access to the application before, during and after the failure. Cisco can provide the most advanced and comprehensive suite of business continuity network solutions, and can be integrated with business continuity technologies developed by leading partners in the industry.
Cisco's flexible and large-capacity network solutions for enterprises can support a complete set of business continuity IT solutions from backup to data center mirroring. Cisco business continuity network solutions have the following characteristics:
A highly sustainable data center IP network;
Extensible intelligent storage network;
Large-capacity, low-latency SAN expansion and data center interconnection;
Persistent and flexible user access.
Below, we will introduce these functions in detail.
Advantages of Cisco Business Continuity Network
These solutions have the following advantages:
Flexibility-the most abundant business continuity network solution, which can meet the different requirements of various enterprise applications for RTO and RPO;
Network availability-practical design, hierarchical hardware and link redundancy, and high software sustainability;
Interoperability with products of leading storage and system manufacturers-low risk, rapid integration and deployment, and low cost;
Low total cost of ownership-enables enterprises to integrate business continuity communication into a flexible network infrastructure, which can reliably support various applications to reduce telecommunications costs;
It adopts industry-leading technologies, including advanced data replication intelligence and multi-protocol support in storage networks, as well as high-density traffic and protocol flexibility in optical fiber networks.
Highly sustainable IP network infrastructure for data center.
In order to ensure the continuous access of applications and servers, and the interconnection between servers in the data center, it is necessary to design and deploy the IP network infrastructure of the data center. Network availability depends on many aspects of the design: redundant modular components in switches and routers, redundant links between switches, routers and servers, and high availability intelligence to switch to standby components, devices and links quickly and transparently. Recovery should also be able to provide advanced services, such as load balancing and security of servers and applications. Data center employees should follow operational best practices, tools and support to quickly respond to technical problems and prevent human failure.
Powerful storage network that is easy to expand.
Storage network is an important part of business continuity strategy. It can separate storage from servers and create a storage "pool" that can be accessed by multiple servers, thus improving data availability. This method is more sustainable than direct attached storage, because even if the server fails, the data is still available.
Storage network helps to improve application and data availability, because it can standardize and centrally manage business continuity solutions, such as tape backup, data replication and data mirroring between multiple systems. Using storage network, the integration of tape backup service can be realized through Fibre Channel or Internet Small Computer System Interface (iSCSI) protocol. Compared with the traditional SCSI backup solution, it can shorten the backup and recovery time, because the backup is carried out on a network independent of the main data center LAN. Cisco has adopted a unique storage technology, called Virtual Storage Local Area Network (VSAN), which can integrate scattered storage "islands" into a unified physical network, which can be centrally expanded and managed without affecting the availability or security of logically independent SAN. For example, VSAN can establish a logically independent network for tape backup. The dedicated bandwidth of tape backup does not affect the performance or availability of other time-critical applications using the SAN (Figure 3). Other advantages of Cisco storage networking solutions include the ability to integrate advanced storage intelligence into network segments. This intelligence includes virtualization and data replication software provided by Cisco partners. With this intelligence, storage administrators can centrally control the growth of storage and achieve consistent and transparent data replication and mirroring.
Figure 3 Multilayer Intelligent Storage Network
Cisco provides this powerful intelligent storage network infrastructure based on Cisco MDS 9000 series multilayer switches.
High-capacity and low-latency data center interconnection
At present, many organizations integrate various applications, servers and storage into a few facilities, so it is necessary to ensure that these integrated data centers will not become a single point of failure. According to the pre-determined risks, the standby data center can be located in another office area, another city, another country or anywhere on the earth. In order to support the business continuity solution chosen by enterprises, enterprises need to formulate corresponding interconnection strategies. There are two general deployment schemes: office area/metropolitan data center interconnection and long-distance long-distance connection (Figure 4).
Figure 4 Data Replication and SAN Expansion Mode
Office/Metropolitan Data Center Interconnection
Most plans deploy standby data centers in the same office area or city. Because of its short distance, it can minimize network latency and support any business continuity solution, including synchronous disks and data center mirroring. Many high-capacity technologies, including Dense Wavelength Division Multiplexing (DWDM), SONET, SDH and Gigabit/10 Gigabit Metropolitan Ethernet, have high flexibility and can achieve a balance between cost, management and technical objectives. In addition, enterprises can also set up their own metropolitan optical networks, rent capacity from telecom operators, or adopt hosting services.
DWDM- As the first layer technology, DWDM can increase the available bandwidth of optical fiber economically and effectively. It can support any storage protocol, including IBM optical fiber connection, enterprise system connection (ESCON), IBM Sysplex and coupling link, Fibre Channel and iSCSI. DWDM supports all major LAN protocols and supports data, voice and video transmission. Cisco can provide ultra-high network capacity (32 channels, total throughput of 320Gbps), high-density service convergence and service transparency of various client interfaces for metropolitan DWDM systems through flexible sending and receiving methods, comprehensive service protection methods and required amplification functions.
SONET/SDH-SONET and SDH are well-known and widely used technologies, which can simultaneously transmit data and time division multiplexing (TDM) services through electronic and optical interfaces. Recent enhancements enable Ethernet and DWDM technologies to integrate with SONET/SDH, thus fully demonstrating its flexibility to meet the ever-changing network requirements. SONET/SDH can be used for many services, such as SAN expansion. Cisco is the industry leader of the next generation SONET/SDH solution, which can provide all kinds of services, bandwidths and interfaces on one platform-Cisco ONS 15454. At present, it can support both Fibre Channel and ESCON through SONET/SDH.
Metro Ethernet-As a leading solution provider in providing Gigabit and 10Gigabit Ethernet in metropolitan area network environment, Cisco can connect business continuity solutions based on IP and Ethernet through Ethernet metropolitan area network to realize smooth large-capacity transmission, thus providing operability and manageability for enterprises. Because Cisco supports standard protocols such as Fibre Channel over IP (FCIP), it can implement cost-effective business continuity deployment (Figure 5).
Fig. 5 Selecting the Suitable Metro Ethernet Technology
Remote connection
In order to connect data centers that exceed metropolitan distance (more than 200km), Cisco provides various transmission modes. Because most applications will cause excessive delay due to long-distance transmission, Cisco does not recommend enterprises to adopt synchronous mirroring solution for most applications with a distance of more than 200km, but recommends enterprises to adopt asynchronous replication technology. The best service choices for business continuity applications are SONET/SDH and IP. Cisco ONS 15454 multi-service supply platform transmits replication traffic through SONET/SDH, and can choose various protocols, including FCIP, IP and Fibre Channel. IP WAN service can use IPSec and/or MPLS technology to transmit FCIP and IP replication traffic through secure VPN service.
Cisco uses intelligent network solutions to achieve remote asynchronous replication, thus providing protocol conversion, high availability, quality of service and security, enabling enterprises to deploy remote business continuity solutions reliably, economically and effectively.
Persistent and flexible user access
Sustainable data center network, high-capacity and low-latency network interconnection and powerful storage network are three major components of business continuity network solutions. The fourth part is also very important, its purpose is to ensure users' access to applications by using sustainable and flexible technologies, so as to achieve the purpose of forgiveness. When creating a network infrastructure that can meet the requirements of RAO, network administrators should make contingency plans and logistics plans for local and remote employees, applications and communications. These plans should address the following issues:
Do users or remote sites need access to data center applications and communication services?
Can the user re-establish the connection with the standby site in a timely and appropriate way?
Which fallback access method will be used if the primary access method fails? Can these methods achieve the goals of RAO?
Can the network be rerouted automatically in case of network failure?
Do remote locations need multiple links from different ISPs to connect the data center/standby data center?
If the main work campus is unavailable, do users need specific access functions, such as access from the hotel or home?
These problems can be solved in two ways: transparent redirection and manual reconnection.
Transparent redirection-this technology is invisible to users, and it adopts load balancing, server and "path state sensitive" routing, Internet reuse and advanced routing technology. Load balancing can be performed within or between data centers. Intelligent site selection can find the user's physical location and send the user information to the nearest or fastest responding data center. This technology can detect whether a place is connected to the Internet and determine the "status" of the path and server. If a problem is detected, load balancing technology usually transfers the connection traffic to another data center without the user's knowledge.
Reconnect manually-When there is a network failure at the main site or the application is unavailable, users need to establish a new connection with the application. Cisco is an industry leader in accessing the Internet, and can provide various solutions for enterprises. For manual reconnection, Cisco provides scalable and flexible VPN terminal options in its routers, switches, firewalls and VPN concentrators. Cisco IOS? The software adopts powerful failover and rerouting technology, which can not only recover quickly from any network failure, but also be transparent to users. In order to meet the demand of creating a mobile office quickly, Cisco developed a secure wireless network solution based on IEEE 802. 1 1 standard (Figure 6).
Fig. 6 enables users to perform continuous and balanced access to applications and data.
Migrate to Cisco Business Continuity Network
Cisco recommends that enterprises develop and implement corresponding business continuity network solutions according to the following steps:
Conduct business impact assessment of all enterprise applications, and assess the risk and cost of application interruption;
Work with business, IT and network decision makers to develop business continuity plans to meet application impact assessment requirements;
Deploy corresponding business continuity solutions with Cisco and its partners;
Continuously modify and test plans and solutions to meet changing needs.
Under the guidance of Cisco, enterprises can deploy business continuity network as one of the main components of enterprise data center network architecture. Customers can provide advanced services, design and audit corresponding data center infrastructure and business continuity solutions, so as to maintain continuous normal operation.
Business continuity partnership
Cisco's intelligent networking and storage technologies lay the foundation for business continuity products, which can help data center manufacturers maintain business continuity in the event of failure. Cisco can cooperate with leading manufacturers in various ways to realize the integrated and smooth supply of data centers and business continuity solutions, so that enterprises can not only meet the current needs, but also constantly adjust after development and changes. These partnerships provide data center managers with various resources needed to design, deploy and maintain flexible data centers and effectively achieve their corporate goals.
Cisco: the leading manufacturer in the field of business continuity network
Business continuity is the key to business development. At present, many enterprises are integrating data center resources and pursuing efficiency improvement. Therefore, the business continuity strategy must be integrated into all levels of IT, and it should start with the basic network supporting all data center communications. The network can not only ensure continuous access to key business applications, but also ensure the correct recovery of applications and data through replication and mirroring. No matter what faults or disasters occur, effective designs, templates, best practices and network solutions provided by Cisco for data center managers can help data centers keep their business running normally.
The above content comes from the internet, and I hope it will help everyone.