Service Redundancy Daemon Overview
Introduction to the Service Redundancy Daemon
The service redundancy daemon (srd) provides configurable redundancy across multiple gateways on MX Series routers with MPC. You can configure redundancy based on monitored events, including:
Link down events.
FPC and PIC reboots.
Routing protocol daemon (rpd) aborts and restarts.
Peer gateway events, including requests to acquire or release mastership, or to broadcast warnings.
The srd also enables you to manage stateful sync session synchronization across gateways.
Service Redundancy Daemon Components
The following configurable components control srd processing:
Redundancy Event (RE)—A monitored critical event that triggers the srd to acquire or release mastership for redundancy peers, or to trigger warning-only events, and to add or delete signal routes. Monitored events include interface or link down events, rpd events, and acquire or release mastership events from peers.
Redundancy Policy (RP)—A policy that defines the set of actions taken when a redundancy event occurs. Available actions include acquisition or release of mastership, and addition or deletion of signal routes.
Redundancy Set (RS)—A collection of more or more service sets with a common redundancy policy or policies. A redundancy set applies to two or more system gateways. Only one of the gateways is master and the peer or peers are standby at any time. Redundancy policies define the actions to be taken for an RS when the srd detects a triggering event.
Redundancy Group (RG)—A collection of redundancy sets that defines common peering properties across a set of gateways. Redundancy groups allow for different peering settings across same peers.
In the current implementation, a one-to-one relationship exists between redundancy set and redundancy group.
Signal routes—Static routes that are added or deleted by the srd based on mastership state changes.
Routing Policies—Policies that are configured to advertise routes based on the existence or non-existence of signal routes using the if-route-exists condition.
VRRP (Virtual Router Redundancy Protocol) route tracking—TA standard Junos OS VRRP feature, but optional srd component, that tracks whether aa reachable route exists in the routing table of the routing instance included in the configuration and dynamically changes the priority of the VRRP group based on the reachability of the tracked route, triggering a new master router election. The route to be tracked is the a signal route.
Service Redundancy Daemon Constraints
The following constraints apply to srd processing configurations:
A one-to-one relationship exists between a redundancy set (RS) and a redundancy group (RG). One RS can be part of only one RG.
One redundancy policy (RP) can be part of only one redundancy set (RS), but one redundancy set can have multiple redundancy policies. For example, redundancy set RS1 can include redundancy policies RP1 and RP2. Redundancy policies RP1 and RP2 cannot be included in redundancy sets other than RS1.
One redundancy event (RE) can be part of only one redundancy policy (RP), but one redundancy policy can have multiple redundancy events. For example, redundancy policy RP1 can include redundancy events RE1 and RE2. Redundancy events RE1 and RE2 cannot be included in redundancy policies other than RP1.
One monitored interface or link can be part of only one redundancy event (RE), but one redundancy event can have multiple monitored interfaces.
One service set (SS) can be part of only one redundancy set (RS), but one redundancy set may have multiple service sets.
Service Redundancy Daemon Operation
The srd operates as follows:
The srd runs on the Routing Engine. It continuously monitors configured redundancy events.
When a redundancy event is detected, the srd:
Adds or removes signal routes specified in the redundancy policy.
Switches services to the next preferred standby gateway.
Updates stateful sync roles as needed.
Resulting route changes cause:
The routing policy connected to this route to advertise routes differently.
VRRP to change advertised priorities.
To summarize the switchover process:
A critical event occurs.
srd adds or removes a signal route.
A routing policy advertises routes differently. VRRP changes advertised priorities.
Services switch over to the next preferred standby gateway.
Stateful sync is updated accordingly.
The order of routing priorities must match the order of services mastership.