Skip to content

DR Architecture

Overview

  • Efffective DR/BC costs money all of the time
  • Need extra resource (backup or standby resources)
  • Executing a DR/BC process takes time
  • Cost Trade Off: Faster to Recover - more costilier

Backup and Restore

  • Have backups in place -- in an event of disaster restore them in different site/location
  • Cost effective but will takes longer to restore in case disaster happens


backup-restore

Pilot Light

  • Secondary environment with bare minimum set up for faster recovery
    • auto sync data for Databases which are ready to use
    • provisioned but stopped instances


pilot-light

Warm Standby

  • Standby secondary site running the fully functional version of primary infra but with less resources and size.
    • cheaper instance type
    • lower no. of intances in ASG
    • can be upgraded/updated when failover is required
  • Little costlier than pilot light (cheaper than active/active) but faster to recover in case of disaster

warm-standby

Active/Active

  • Have exact and fully functional version of primary infrastructure running in secondary site/location
  • The most costliest (200%) set-up but does not require recovery during disaster event
  • Since both sites run fully functional apps, they can be load balanced as well
    • provides better performance and improves HA

active-active