Failover

Failover is a high-availability (HA) mechanism that ensures uninterrupted service by automatically transferring operations to a backup system component when a failure occurs.

It can apply to:

• Network links
• Storage controllers
• Server nodes
• Virtual machines
• Applications or services

Failover helps maintain continuous uptime, even during:

• Hardware crashes
• Software errors
• Power outages
• Maintenance tasks

• Active-Passive Failover: A primary system actively handles traffic while a secondary remains on standby. When failure is detected, services are redirected to the passive node.
• Active-Active Failover: Both nodes are operational and share workloads. If one fails, the other takes over 100% of the load—ideal for balanced performance and resilience.
• Manual Failover: Requires administrator intervention to redirect services. Less automated but useful in controlled environments or for staged recovery.
• Automated Failover: Uses monitoring and heartbeat mechanisms to detect failure and trigger a switchover without manual input—minimizing downtime risk.
• Application-Level Failover: Specific services (e.g., databases, VMs) are restarted or migrated to alternate systems in response to localized crashes or overload.

• Minimizes service disruption: Users experience little to no downtime during failures, keeping applications and systems operational during critical periods.
• Protects business continuity: Ensures access to data and systems even during maintenance, disasters, or infrastructure failures—crucial for regulated industries.
• Reduces support response time: Automated failover eliminates the delay of manual troubleshooting, allowing IT teams to focus on root cause analysis later.
• Enables maintenance with zero downtime: Systems can be patched or updated while services are temporarily rerouted to backup nodes or locations.
• Strengthens disaster recovery strategies: Failover complements backups and replication, enabling near-instant access to redundant resources after a system event.

Open-E JovianDSS includes built-in high-availability clustering features that provide enterprise-class failover for storage systems:

• Active-Passive and Active-Active clustering: Configurations allow for fully redundant nodes that automatically take over services and mount points when a failure is detected.
• Heartbeat and watchdog monitoring: Nodes communicate constantly to verify health. If a heartbeat is lost, failover is triggered immediately to the standby node.
• Shared storage with synchronized metadata: Both nodes have access to the same storage pool, ensuring seamless transition and up-to-date data visibility.
• Failover of NAS, iSCSI, and Fibre Channel targets: Not just files, but block storage targets and protocols are fully supported in HA setups.
• Email alerts and logging: All failover actions are logged and reported, allowing for later auditing and diagnostics.

• Test failover procedures regularly: Simulate failures under controlled conditions to confirm automatic switchover works as expected—and to train staff.
• Monitor HA components proactively: Use SNMP, logs, and dashboards to detect degraded hardware or link errors before failover is triggered unnecessarily.
• Use redundant power, cooling, and networking: Failover is most effective when supporting infrastructure also avoids single points of failure.
• Keep firmware and OS updated: Prevent known bugs from interfering with cluster operation or HA logic by following a strict update policy.
• Document roles and dependencies clearly: Ensure that all systems know which node handles what service—and how data is synchronized between them.