No public concern: AWS lovers (iloveaws)
text | silent devil (please indicate the source)
Website: www.iloveaws.cn
In Beijing, between 2019 Nian 6 Yue 2 Ri 2:00 AM to 13:48 PM, with multiple fiber Waduan in road work in the area last night's CN-NORTH-1 (Beijing China region), resulting in the first available area region can not both EC2 instance EC2 instance in the region of the entire new CN-NORTH-1 in.
This is a major accident AWS region of China, but also to remind us, take the initiative, to learn the knowledge of disaster recovery, disaster recovery plan as soon as possible to do business is very important; not a cloud provider can guarantee lasting no problem, when a disaster occurs , rational use of DR plans, business recovery in the first time, the impact of the disaster to minimize business impact is our goal.
In front of us the details about Disaster Recovery (DR) include:
-Disaster the definition of content and Recovery (DR) Disaster Recovery
- Recovery Time Objective (RTO) and recovery point objectives (RPO)
- and disaster recovery (DR) capabilities and related AWS Service-1
- and disaster recovery (DR) related the functions and services AWS -2
- and disaster recovery (DR) relating to service functions and AWS -3
-AWS example disaster recovery scenarios - (1) backup and restore
-AWS disaster recovery scenarios example - (2) used in AWS Pilot Light fast recovery
Today, we continue Disaster Recovery (DR) content, we began to introduce hot standby policy third strategy -AWS four kinds DR strategy.
- Strategy 3, AWS Redundancy -
Hot Standby solution is introduced in front of our strategy of Pilot Light extension (assembly and preparation).
The term "hot standby" is used to describe DR strategy: to create a scaled-down version of the full functionality of your business operating environment is always running in the cloud, when a disaster occurs fast switching.
Because the hot standby policy will create a set of service your business is always running in the cloud, thus further reducing the recovery time in case of disaster.
By identifying your core business systems, you can completely replicate these systems on AWS and it is always running in the cloud, these servers can start on the Amazon EC2 instance is the smallest size to the smallest possible size of the operation, it does not require we need a carrying capacity of normal business, but to have all of the features, such as it can be used for non-production work, such as testing, quality assurance and internal use.
During DR, minimizing environment we created earlier to be able to quickly scale to handle the production load. In the AWS, by adding more instances to the load balancer or adjust the size of the complete example of rapid expansion.
Preparation Phase
The following figure shows the Hot Standby solution preparation stage, where traditional data center environment and the environment AWS run in parallel.
(Using traditional data center and AWS way of example only, of course, also applicable to multi-zone AWS, cloudy environment)
The key steps to prepare:
Amazon EC2 instance set to replicate or mirror data.
Creating and maintaining AMI.
With a minimum of Amazon EC2 instance and AWS infrastructure to run the application.
Depending on your production environment patching and updating software and configuration files.
Recovery phase
In the case of failure of the production environment, the environment will be scaled hot backup for the production of a load, and change the DNS record all traffic routed to the AWS.
The key steps to recover:
- increasing the load balancer Amazon EC2 number (horizontal expansion)
- start the application (vertical extension) over a larger Amazon EC2 instance type based on business needs.
- Manually change DNS records, or use the Amazon Route 53 automatic health check, so all traffic will be routed to the AWS environment.
- Consider using Auto Scaling to adjust the size or accommodate the increased load.
- Add elasticity or extend the database.
AWS Disaster Recovery program example above - Content (3) AWS hot backup strategy is over, we see you next time!
(to be continued)
No public concern: AWS lovers (iloveaws)
text | silent devil (please indicate the source)
Website: www.iloveaws.cn