Updated on 2022-11-16 GMT+08:00

Service Resilience

DEW implements fault isolation, data backup, and traffic control to improve service resilience and enhance user data security.

Fault Isolation

  • The inter-region isolation design of DEW ensures that the faults in a region do not affect the DEW services in other regions.
  • DEW servers and HSMs adopt the AZ-level DR design, so that the faults in an AZ do not affect DEW availability. In the case of a fault, DEW automatically shields the faulty AZ and switches traffic over to other another AZ, smoothly scheduling workloads.
  • DEW servers and HSMs are deployed in cluster mode. If any single-server or single-HSM fault does not affect DEW availability.

Data Backup

DEW keys are replicated among multiple HSMs to avoid permanent key loss in the case of an HSM fault. DEW data (non-sensitive data) is replicated among multiple servers and database instances, and backed up in real time to prevent data loss.

Flow Control

DEW can meet the SLA target of 99.95% availability and provide a large quota of API calls for each user. If a user has used up their quota of API calls, DEW will restrict their subsequent API calls to ensure service availability.