VM High Availability (HA) — Complete Setup Guide & How It Works
  

George Fady Lv2Posted 2026-May-30 06:48

Last edited by George Fady 2026-May-30 06:54.

Last edited by George Fady 2026-May-30 06:52.

High Availability (HA) in Sangfor HCI 6.9.11. What is VM HA?[td]
ItemDescription
HA MeaningHigh Availability
PurposeAutomatically restart VMs on another healthy node if the current node fails
BenefitsReduces downtime and keeps services running
Common Failure TypesHardware failure, network outage, node crash

2. Requirements for HA[td]
RequirementDetails
Shared StorageVM storage must be on shared storage accessible by all nodes
Supported StorageaSAN, iSCSI, Fibre Channel (FC)
Unsupported StorageLocal storage
Run LocationMust be set to Automatic Selection
HA SettingHA must be enabled manually on the VM
Important Note[td]
Cluster TypeBehavior
2-Node ClusterHA will NOT trigger if the storage network interface disconnects
3+ Node ClusterFull protection supported

3. How to Enable HA[td]
StepAction
1Go to Compute
2Select the VM
3Click Edit
4Enable High Availability
5Set Run on Node to Automatic Selection
6Save the settings
VM Placement Behavior[td]
SituationResult
VM StartupHCI selects the least-loaded node
Node FailureVM automatically restarts on another healthy node

4. HA Trigger Conditions[td]
Trigger TypeDescription
Management Network FailureHost becomes unreachable
Physical Network FailureVM loses physical network connection
Storage FailureVM storage path becomes unavailable

5. HA and VM Priority[td]
Priority TypeBehavior
High Priority VMResources are pre-allocated
Memory HandlingNo memory reclaiming or ballooning
During FailoverHigh Priority VMs restart before normal VMs

6. Resource Reservation for HA[td]
FeatureDescription
CPU ReservationReserves CPU resources for HA
Memory ReservationReserves RAM resources for failover
BenefitGuarantees enough resources during failover
Scheduling ImpactReserved resources are excluded from normal scheduling
Tip[td]
Recommendation
Configure resource reservation under Reliability > Resource Scheduling when using many HA-enabled VMs

7. HA vs Live Migration vs Resource Scheduling[td]
FeatureHALive MigrationResource Scheduling
TriggerNode failure / network lossManual or automatic by adminCPU/RAM threshold exceeded
VM StateVM restarts on another nodeVM moves without downtimeVM moves without downtime
DowntimeBrief restart downtimeZero downtimeZero downtime
RequirementsShared storageShared storage + networkShared storage + policy

Like this topic? Like it or reward the author.

Creating a topic earns you 5 coins. A featured or excellent topic earns you more coins. What is Coin?

Enter your mobile phone number and company name for better service. Go