Disaster Recovery Colocation

Disaster recovery colocation means renting rack space, power, cooling and connectivity in a secure third-party data centre and using it as the recovery site for your systems. Your data is replicated there continuously, so if the primary site fails you can fail over and keep running. You own the servers; the facility provides the resilient building, power and network.

Set RTO and RPO for each workload tier, choose a colocation site far enough from the primary to avoid shared risks, replicate servers and virtual machines to it, keep immutable backups, write failover runbooks and test them. Layots recommends a full failover test twice a year and a partial test each quarter, because untested DR is the most common cause of failed recovery.

Both work. Colocation suits steady, regulated or latency-sensitive systems and organisations that want to own the hardware. Cloud suits recovery capacity you only pay for when you fail over, with automated failover to cloud nodes in under 15 minutes. Many designs combine the two, and Layots chooses per workload.

RTO is how long you can be down; RPO is how much data you can afford to lose. Most enterprise line-of-business systems target an RTO under four hours and an RPO under fifteen minutes, while transactional or financial systems often need near-zero RPO through synchronous replication.

Tier-3 and tier-4 facilities including Yotta, STT and Equinix, with managed rack, power, cross-connect and remote-hands services.

A full failover test twice a year and a partial tabletop or component-level test each quarter. Layots schedules and runs these tests and issues a report with the actual measured RTO against the contracted target.