edit-icon download-icon

Architecture of replica set

Last Updated: Jan 17, 2018

ApsaraDB for MongoDB automatically creates a three-node replica set for you to use. You can directly operate on one primary and one secondary node. The following figure shows the system architecture:


HA Control System: Instance highly available detection modules are used to detect and monitor the operating status of MongoDB instances. If the system determines that the primary node instance is unavailable, it switches over to the standby node, to guarantee the high availability of MongoDB instances.

Log Collection: This process collects MongoDB operating condition logs, including instance slow query logs and RAM logs.

Monitoring System: This system collects MongoDB instance performance monitoring information. This currently includes basic metrics, disk capacity, network requests, operation counts, and other core information.

Online Migration System: When the physical server that runs an instance fails, the online migration system re-creates an instance based on the backup files in the backup system. This guarantees that the business is not affected.

Backup System: This system backs up MongoDB instances and stores the generated backup files on the OSS system. Now, the MongoDB backup system allows users to customize the backup settings and temporary backup configuration. Files are retained for 7 days.

Task Control: ApsaraDB for MongoDB instances support various management and control tasks, including instance creation, configuration changes, and instance backup. The task system flexibly controls tasks and executes task tracking and error management based on the commands you give.

Thank you! We've received your feedback.