Disaster Recovery & Failover

Replication interval is the whole conversation

Every meaningful number in a disaster-recovery design follows from two things: how often workloads replicate, and how quickly people can reach them once they run somewhere else. A fifteen-minute interval and a four-hour interval cost differently and lose differently, and no amount of platform branding changes that arithmetic.

The second half is network. If a plant's production line depends on a server that has failed over eighty miles away, the circuit, the firewall rules and the DNS changes decide whether people can actually work. We design that path at the same time as the replication, because a workload running in Carbondale that nobody can reach is not a recovery.

• Replication interval set per workload and stated plainly in the design
• Access path (circuit, VPN, DNS, firewall) planned before failover, not during
• Recovery order and declaration authority written into the plan
• Failback treated as a scheduled project rather than an afterthought

The platform behind replication

Replication into our recovery infrastructure is built primarily on Veeam Backup & Replication, which is also our main backup platform for larger and more complex virtualized environments, particularly Hyper-V clusters. Using one platform for both jobs keeps the backup chain and the replica of a virtual machine understood by the same engineers and covered by the same testing.

Where an environment is protected by Cove Data Protection by N-able or by a Datto appliance, those platforms provide local and offsite backup and recovery, not replication into the Mayer Networks datacenter. That is the architectural line worth understanding when comparing designs, and it is why a continuity requirement usually points toward Veeam.

• Cove: local recovery storage plus protected offsite cloud backup
• Datto: local recovery appliance plus protected offsite cloud backup
• Veeam: enterprise backup, virtual machine replication and optional failover into Mayer Networks recovery infrastructure

Where this sits in the three-layer model

Production is your active environment, onsite or in cloud services. Backup is protected recovery copies of the data, usually a local copy for speed and an offsite copy for separation. Disaster recovery is replication of critical systems to Mayer Networks infrastructure so they can be recovered or failed over if production infrastructure becomes unavailable.

The layers overlap in purpose and not in mechanism. Backups are deliberately kept away from production so an attacker cannot reach both; replicated workloads are deliberately kept close in time to production so operations resume quickly. Designing one to do the other's job is how organizations end up disappointed at the worst moment.