Create a data recovery plans on the cloud is a very important point to keep track infrastructure, DR needs and possible failover duration.
Public cloud IT department to provide an excellent opportunity to achieve business continuity / disaster recovery plans, without having to spend huge sums of money to build exclusive data center. With cloud data after recovery system, the cloud can be used as basic data store or even run the application at the time when the main system problems.
When building a DR plan, the first step is to look for application delivery IT services and decide what needs to be protected when a disaster occurs. List of applications and services, this means creating the need to run many companies have turned to virtualization as its core server deployment model; however, still need to consider the physical cloud server complete data recovery plan should include the following:
Delivery infrastructure for physical and virtual servers. These include Active Directory (AD) server, DNS / DNCP servers and applications.
- For the delivery of the physical server applications on a physical server is also why the delivery service, which requires a little better reasons; this may include expansion and performance requirements, or use custom hardware and operating systems, however, may be able to cloud recovery services. Some of these components help virtualization.
- Used to deliver virtual server applications may have dozens or hundreds of virtual machine used to implement each application needs to identify and record, view memory, virtual memory and processor requirements...
The best server infrastructure is determined in advance, because when a disaster occurs the first time these systems require recovery services can be pre-configured in the cloud running AD, DNS and DNCP services, and synchronization, and their internal instance, make the process easier for DR It can also be achieved more quickly.
. To get on the cloud DR to work successfully, the network configuration is essential to understand this means that it takes time to understand the interdependence between the application layer network, including security and firewall configuration cloud data recovery related issues:
- Is there another application or server delay between dependent?
- Is there a firewall rule to East-West traffic within the station?
- What external customer-facing applications bandwidth requirements are?
Determining cloud data recovery needs
Assumed disaster event occurs, each application requires immediate recovery, and this is not very practical. Instead, it should be based on a series of conditions to distinguish the priority application, to determine how quickly, and what needs to be restored and data synchronization system operators in determining the level of service recovery applications, you can use some criteria to measure:
. Recovery time objective measure before it and run the backup program should be able to tolerate a long time off the assembly line; usually measured in minutes or hours, for example, dropped zero RTO represents totally unacceptable, and a one-hour RTO means that the application must be complete recovery occurred within an hour DR.
Recovery point objective. It measures how much data loss can be tolerated once the application is running again. RPO of zero means that all data must be restored to the point of disaster, and after a 24-hour RTO means to recover data may be outdated or system for 24 hours.
Service Level Targets .SLO measure the recovery of the overall application. For example, the agreement may be used in the recovery of 90% within four hours. The more stringent requirements SLO more infrastructure support and may require more manpower to achieve, thus leaving there is a certain degree of flexibility to help manage DR costs.
SLO allows to distinguish between data and application priority. For example, an online credit card processing systems require very low zero RPO and RTO. Expect such a system will never lose information is reasonable. At the other extreme is responsible for reporting applications may be able to tolerate 24-48 hours of data expiration time, because its data is extracted from other applications inside out. most other systems fall between these two extremes.
Establish the correct data recovery needs including cloud applications and business owners to communicate, because they understand the importance of the application from the point of view of experience, business owners would think that all of its applications are very important - they understand until needed for recovery up costs, so we can tell them to assess the cost of different options.
The last point is the level of service: some of the stringent requirements, such as zero-PRO, cloud-based DR can not be achieved, because there will be delays between local and cloud location of these applications need to be excluded from the cloud-based DR,. and to provide more customized products DR.
DR service will run long?
A final question is, how long the service will run in the public cloud make such decisions depend on the type of event occurred, not all disasters will lead to the collapse of all the online functions will continue to exist some marginal types of events, such as...:
Server is lost. Either physical servers or virtual server host. Lost virtual server can be severe, but it may not be serious, the application needs to run instead in DR mode.
Multi-system missing. For example, if the shared storage array problems, it could lose more than one application.
Data center is lost. In the worst case, the entire data center is lost, or can not visit all the services needed to run in DR mode.
Sometimes, the service needs to be moved a few hours or days. When the entire site are lost, the demand may be weeks or months to run a DR service until the equipment before reconstruction cloud recovery services will be used to service activities billing, so when selecting DR service it is very important assessment point.
www.get-top-news.com Get Top News, Latest News and Current News from get-top-news.com. Latest Current News: U.S., China, World, Entertainment, Health, Finance, Business, echnology, Travel, Politics, Sports.Traditional Chinese: 雲數據恢復：文檔是成功的關鍵