High availability solution of the edge cloud

Disaster recovery for multiple compute instances

To ensure local high availability, your business should have the ability to handle heavy loads and single points of failure (SPOFs) that can cause service interruptions.

1. Traffic distribution:

   You can deploy your business on multiple ENS instances and use ELB to distribute traffic across the ENS instances.

2. Health check:

   1. ELB performs health checks to determine the availability of ENS instances. The health check feature ensures business availability even if some backend ENS instances are abnormal.

   2. After you enable the health check feature, ELB distributes new requests to healthy ENS instances whenever unhealthy instances are detected. After the unhealthy instances recover, ELB automatically resumes sending requests to the instances.

   
   Note

   For more information about ELB, see ELB.

3. High availability software:

   1. In traditional data centers, servers can announce their IP addresses over ARP. Then, the servers can provide services through the IP addresses. This capability is required by common applications in many scenarios. For example, Keepalived and Heartbeat can be used to keep IP addresses unchanged in disaster recovery scenarios.

   2. You can use HAVIPs to implement the same deployment solution as a traditional data center. This allows you to implement IP address failover in a high-availability architecture that contains multiple servers in the same node, and ensures that the private IP addresses for external services remain unchanged.

   
   Note

   For more information about HAVIPs, see What is an HAVIP?

Data disaster recovery

Edge Node Service provides storage such as disks, NAS, and EOS for data storage.

1. Multi-replica data redundancy:

   1. Within a node, data stored on disks and NAS is distributed in three replicas in the storage cluster. This ensures high availability and reliability of at least 99.9999%.

   2. EOS uses the data redundancy mechanism based on erasure coding to ensure data durability and availability.

2. Data backup:

   You can use snapshots of Edge Node Service to back up local disk data. When you create a snapshot, the current state of data blocks on a disk is captured. You can use snapshots to restore data, build development and testing environments, or create custom images for batch deployment of your business.

   

Cross-node network high availability

1. ENA can provide enterprises with fast connections, such as accelerated connections between edge nodes, accelerated connections between data centers, accelerated connections between the internal network and Alibaba Cloud central cloud, and accelerated connections between different types of public clouds.

2. With ENA, you can connect virtual private clouds (VPCs) in different regions or on different edge nodes. This helps build a high-speed and secure data transmission channel for business systems to communicate with each other over the internal network.

3. For more information about ENA, visit the ENA product page.

Cross-node service high availability

1. Multi-node deployment of your business system:

   You can deploy a business system on two or more edge nodes to implement multi-node disaster recovery for the business system.

2. Cross-node high availability: You can use Global Traffic Manager (GTM) provided by Alibaba Cloud DNS to achieve high availability. You can configure the following items on GTM to build a high-availability environment for external services:

   1. Configure a domain name: the domain name of the business system that provides external services.

   2. Configure address pools: the public addresses for external services, which are EIPs in ENS for external services. You can add all addresses for external services on each node to different address pools.

   3. Configure load balancing policies for address pools: load balancing policies for all addresses in the address pools, such as returning addresses by weight.

   4. Configure the address working mode: the working mode for addresses in the address pools, such as Intelligently Returned and Always Online.

   5. Configure health check for address pools: health check policy for addresses, such as check protocol and check port.

   6. Configure access polices: including intelligent DNS resolution, primary and secondary address pool sets, and policy for switchover between address pool sets.

      1. You can configure a primary/secondary address pool to implement cross-node primary/secondary disaster recovery.

      2. You can configure all address pools are configured as available address pools to implement multi-active and high-availability services.

      Note

      GTM provides various configuration templates to simplify the configuration process, such as Primary/Secondary Disaster Recovery and Multi-active Load Balancing. You can use these configuration templates to help you set up a high-availability environment quickly. For more information, see Global Traffic Manager3.0.

3. Cross-node high availability of internal services: You can deploy DNS for Multicloud Integration in private mode to implement cross-node high availability of internal services. DNS for Multicloud Integration provides all-in-one intelligent resolution services for internal networks, including intelligent DNS resolution of internal and external domain names, disaster recovery and scheduling of primary/secondary data centers, and unified management of Alibaba Cloud DNS. DNS for Multicloud Integration can replace common open-source DNS services.

   1. High availability of the internal DNS service: DNS for Multicloud Integration supports high-availability deployment to implement cross-node primary/secondary disaster recovery. If your business is normal, the primary DNS is used. If the primary DNS fails, the cross-node secondary DNS is used.

   2. High availability of internal services: Internal services are accessed by using domain names. You can configure domain name resolution records and policies in DNS for Multicloud Integration. This implements multi-active and primary/secondary deployment of internal services.

4. Cross-node disaster recovery for business data

   1. High availability of images: The image service of the edge cloud stores system images in the central cloud of Alibaba Cloud. This allows you to use the capabilities of the central cloud to ensure high availability and high reliability of system images in the edge cloud. If you want a system image to be available on all nodes in the edge cloud, you can create a custom image from the ENS instance. When you want to deploy the same service on other nodes, you can pull the image from the central cloud to create an ENS instance.

      Note

      For more information about the image service, see Images.

   2. Cross-node disaster recovery for data: As mentioned before, you can create a snapshot of a disk to back up data on a node. The edge cloud also provides cross-node snapshot replication, which allows you to replicate snapshots to other nodes. If an edge cloud node fails, you can restore your business and data on other nodes.

   3. Disaster recovery for databases: The edge cloud does not provide native database services. You need to deploy your own database service. You can back up data across nodes by using the built-in data disaster recovery mechanism of the database service, or use snapshots to back up data to other edge nodes.