Continuity and Resilience

Disaster Recovery & High Availability

Recovery you have proven, not recovery you have promised.

Disaster Recovery & High Availability is an application-level continuity platform. It recovers the application itself — namespaces, workloads, stateful sets, services, configuration, secrets and persistent volumes — in dependency order, to a running and verified state. Conventional tools restore machines and leave the operations team to work out what to start, and in what order, while the incident is still running.

Downtime in a regulated organisation is not only lost revenue; it is a supervisory event. Supervisors increasingly ask for documented recovery objectives that have been tested in practice, and for an audit trail showing they hold. Neor gives you objectives you can configure, rehearse and evidence, on infrastructure you own and inside a boundary you control.

Architecture

How it is put together

Recovery Orchestration

Namespace-level restore with label-selector filtering, so a single failed service can be brought back without touching the workloads beside it.

Failover Engine

Continuous health polling of every registered cluster; once a cluster stays unhealthy beyond its grace period, the engine taints it, drains pods through graceful eviction and reschedules them onto healthy capacity.

Global Control Plane

A management cluster you host yourself, holding cross-site scheduling, failover policy and health state for the whole estate.

Member Clusters

Production, standby and dedicated recovery clusters, in your own data centres or hosted environments, register as managed members, with workloads placed by weight or by available capacity.

Backup and State Store

Scheduled full and incremental snapshots of application namespaces and cluster state, written to the storage backend you nominate: S3-compatible object storage, on-premises NFS, or a private object store.

Capabilities

Application-Level Recovery

Recovers namespaces, deployments, stateful sets, services, config maps, secrets and persistent volumes in dependency order to a running application state, not a machine image you still have to reassemble.

Automated Detection and Failover

Cluster health is monitored continuously; when a cluster crosses its unhealthy threshold, eviction and rescheduling begin automatically, without waiting for an operator to intervene.

Cross-Site and Cross-Cloud Failover

Failover runs between clusters in one site, between separate data centres, and between on-premises and hosted environments, including geo-distributed deployments across availability zones and regions.

Active-Active and Active-Standby

Both topologies are supported under the same control plane, with continuous health monitoring and configurable eviction tolerances.

Configurable, Auditable Objectives

Monitoring intervals, unhealthy grace periods and eviction timeouts are set explicitly, so RPO and RTO become engineering parameters you can read, change and report rather than numbers in a policy document.

Rehearsed, Evidenced Recovery

A drill exercises the same automated path a real incident would take, and every run leaves an auditable record, so recovery is demonstrated rather than asserted.

Surgical and Full Restore

Namespace-level restore handles a single service failing in production; full snapshot restore handles the loss of an entire cluster, limiting blast radius in the first case and rebuild time in the second.

Sovereign Orchestration

The platform runs entirely inside your environment, with no call-home requirement, so failover decisions, recovery playbooks and backup data never cross a boundary you do not control.

Where it fits

A core transaction platform has to survive the loss of an entire data centre, and the board wants the switchover time on record.
An auditor asks for proof that the recovery plan was exercised this year, not merely written and filed.
One service fails in production overnight and must be restored on its own, without disturbing the workloads running alongside it.
Applications are split between an on-premises site and a hosted region, and failover today means two separate manual runbooks that have never been run together.
A public-sector system must keep every copy of its data inside national borders while still holding a second live site for continuity.

What you end up with

A continuity capability you own outright: recovery objectives that are configured, automated, rehearsed and evidenced, running on your own infrastructure and inside your own jurisdiction.