Managing Stateful Line Module Switchover; 1:1 Redundancy Model; Seamless Preservation Of Subscriber Sessions - Juniper JUNOSE SOFTWARE FOR E SERIES 11.3.X - SERVICE AVAILABILITY CONFIGURATION GUIDE 2010-10-08 Configuration Manual

1:1 Redundancy Model

Seamless Preservation of Subscriber Sessions

Related
Documentation
Copyright © 2010, Juniper Networks, Inc.
NOTICE 01/27/2004 19:54:06 ha (): High Availability disabled due to secondary line card
in slot 9 down
The commands used to configure stateful switchover for SRP modules and line modules
are very similar. The configuration modes are the same. The keywords to configure stateful
switchover for SRP modules and for line modules are different. When redundancy mode
is configured for high availability on E120 and E320 routers with dual SRP modules, high
availability is enabled only for the corresponding SRP modules. You must explicitly specify
high availability for the line modules to enable stateful switchover of the line modules
to occur.
Line module high availability uses a 1:1 redundancy model to maintain subscriber sessions,
during and after a switchover of the primary line module to the secondary line module.
This feature is supported only on E120 and E320 routers installed with ES2 4G LMs and
Service IOAs. You can enable line module high availability feature only on the compatible
line modules and IOA combinations. On routers in which line module high availability is
disabled or not available for configuration (cold-restart support is not present), all
subscriber sessions are forcibly terminated when a line module fails or stops responding.
This mechanism is known as a cold-restart.
When the secondary line module in the HA configuration becomes the primary line module,
all the applications on it recover to a stable state to operate normally in the place of the
failed line module. The new primary line module maintains existing subscriber sessions
and continues to forward subscriber data traffic after the switchover. A brief subscriber
data disconnection occurs for two minutes during the switchover. Diagnostic functions
are still run on the failed line module to ensure that the hardware on the defective line
module is still usable. The exception might have been triggered by a hardware fault that
the diagnostic services might discover using its testing mechanism. For the stateful line
module switchover to work correctly, the current subscriber information must be made
accessible to the newly configured primary line module. The states of subscriber sessions
that were on the failed line module are mirrored to the secondary line module, which has
taken over as the primary controller.
Only those subscriber sessions whose operational status is up are guaranteed to be
retained. Other subscriber connections that are in the process of fluctuating between up
and down states might not be preserved. Such a technique of not retaining the subscriber
sessions that are constantly alternating is adopted because of the possibility of occurrence
of faults during transitional periods of states. If subscriber sessions in these transitional
states are not preserved, the possibility of the same problem occurring on the secondary
line module after the switchover is reduced.
Stateful Line Module Switchover Overview on page 68
Stateful Line Module Switchover Platform Considerations on page 70
System Operations When Stateful Line Module Switchover Is Enabled on page 74
Chapter 4: Managing Stateful Line Module Switchover
69