Chassis Cluster Resiliency
This topic explains how Junos OS uses a layered resiliency model to provide resiliency on Firewalls operating in a chassis cluster configuration. In the event of a software or hardware component failure, the layered model helps ensure that system performance and service continuity are maintained.
Layer 1 To Detect Hardware Faults and Software Failures
Layer 1 is responsible for identifying and detecting components that cause software failures or negatively impact system performance. When such a condition is detected, the system generates notifications in the form of alarms, syslog messages, or SNMP traps to alert administrators of the failure.
Layer 2 To Probe Critical Paths
Layer 2 monitors the system’s critical data paths to detect hardware and software failures that are not identified by Layer 1.
Heartbeat communications verify the operational state of the path between two endpoints. If any component along the path fails, communication is disrupted, and the system's health status is conveyed through heartbeat messages exchanged between the endpoints.
Layer 3 to Detect Control Link and Fabric Link Failure
Layer 3 evaluates system health information from Layer 1 and Layer 2, exchanges health status between the two nodes over the control links and fabric links, and makes failover decisions based on the health of both nodes and the heartbeat status of these links. When a failure is detected, the system generates notifications in the form of alarms, syslog messages, or SNMP traps.
Layer 3 addresses the following software-related conditions:
-
em0 interface flapping
-
Control path hardware or software component failure.
-
Fabric link failure while the control link remains operational.
-
Control link failure while the fabric link remains operational.
-
Simultaneous failure of both the control link and the fabric link.
Benefits
-
Improve failover time and system stability.
-
Identify the precise location of the fault or failure.