Route heath injection ?
Route Health Injection (RHI) delegates the control of routing protocol announcement to a server based on the health of a service and the connectivity of the server to the network :
- If the service is alive, a /32 route is injected into the network using a dynamic routing protocol announcement.
- When the service goes down because of a software issue, the /32 host route is withdrawn by the server from the network using a dynamic routing protocol triggered update.
- If the server loses it’s connectivity to the network, the upstream router automatically withdraws all the /32 host routes learned from this disconnected peer.
Route Health Injection for IP Anycast deployment
Route Health Injection can be used to advertise the same IP address from different locations geographically dispersed. User is directed to the nearest server from the network perspective. Strong resiliency is achieved for critical services in this IP Anycast.
Route Health Injection in a Layer 3 Citrix ADCHA pair
Citrix ADC High Availability is usually created by pairing two netscaler nodes located in the same management subnet and sharing the same vlans. The primary node hosts the active SNIP and VIP addresses, the secondary node is synchronized with the primary node. If the primary node goes down or loses it’s connectivity, the SNIP and VIP addresses move to the secondary node which gets primary. Both Netscaler nodes share the same network configuration and network resiliency is achived by using first hop reduncdancy protocols (HSRP, HSRP vPC, VRRP, …), vPC, VSS or multi chassis etherchannel, and firewall clustering.
Failure or flapping of one of the redundancy mechanism breaks network connectivity.
In a layer 3 Citrix ADC HA deployment, each ADC instance is deployed in a separate subnet with independant internet connectivity, routers, and firewalls. All those equipments are located in separate network rooms. Connectivity in the Citrix ADC HA pair and between ADC and users / backend servers are routed and rely on dynamic routing protocols.
Dynamic IP routing in Citrix ADC
Citrix ADC support both dynamic and static routing. The main objective of running dynamic routing protocols is to enable route health injection (RHI), to enable upstream network infrastructure to choose the best path among multiple routes to a reach a geographically distributed virtual server.
Multiple routing tables in Citrix ADC
Citrix ADC embeds 3 different routing tables :
- NS Kernel routing table : Used by Citrix ADC in packet forwarding. Holds subnet routes corresponding to the NSIP and to each SNIP and MIP. Holds any static routes added through the CLI. This routing table is configured through the GUI or CLI interface.
- Network service module : Contains the advertisable routes distributed by the dynamic routing protocols to their peers in the network. This routing table is configured using the vtysh mode (ZebOS).
- FreeBSD routing table : Facilitate initiation and termination of management traffic (telnet, ssh, etc.).
Citrix ADC routing suite based on ZebOS® IP Infusion Layer 2, Layer 3, MPLS network stack. ZebOS® can be accessed using vtysh netscaler CLI command. IPv4 & IPv6 supported dynamic routing protocols are :
- Routing Information Protocol (RIP) : version 2 & RIPng for IPv6
- Open Shortest Path First (OSPF) : version 2 & version 3 for IPv6
- Border Gateway Protocol (BGP)
- ISIS Protocol
Routing tables synchronization
NS kernel routing table (GUI & CLI configuration) and Network service module (ZebOS) are ynchronized using the following vtysh commands :
- Use ns-route-install commands to push routes from ZebOS to NS Kernel routing table
- Use redistribute kernel command in ZebOS dynamic routing protocols configuration to advertise routes configured in NS Kernel routing table.