hi human interface®
hi human interface®
English
Breadcrumbs

High Availability

High Availability

Overview

A hi installation can run on one server or use the standard three-node cluster design. Kubernetes supervises the application services in both deployment models.

High availability is a property of the complete commissioned design. It depends on server, storage, database, network, client-access, backup, monitoring, and recovery decisions. The application architecture alone does not define an availability target, recovery time, or data-loss objective.

Service Architecture

Single-Server Deployment

Single Server Deployment

A single-server deployment is not redundant. All hi services run on the same machine, and a failure of that machine stops the hi system.

Multi-Server Deployment

Cluster Deployment

When redundancy is required, the standard multi-server hi deployment uses three cluster nodes with the normal MongoDB topology and storage arrangement. Those database and storage components are part of the standard three-node design, not project-specific choices. Confirm server placement, networking, client access, and failure tolerance against the commissioned site design.

The hi web interface presents the installation as one logical hi system. Use the hostname or address supplied during commissioning.

Service Supervision

Kubernetes can restart and place application services according to the commissioned deployment configuration. A service restart can temporarily affect the function provided by that service while it initializes and reconnects.

See System Status for the Services and Node or Cluster indicators shown in the hi web interface.

Operational Implications

Service, server, database, storage, and network failures can have different operational effects. Cluster maintenance and infrastructure recovery normally require specialist support.

Warning

Cluster failures and recovery actions can critically affect system availability and database integrity, with a risk of data loss. Contact Riedel Customer Success through Customer Support before infrastructure intervention.

Limit customer checks to the status information available through normal authorized access. Recovery and failure testing require specialist guidance and the commissioned procedure. A green status indicator alone does not establish service or data recovery.

Maintain verified off-system Backups even when the deployment includes redundant infrastructure. See Configurations for the difference between Snapshots and Backups.