EE connection or RTP pipe fails to activate when using policy-based routing for EE traffic
When policy-based routing is defined with multiple policy routing rules to separate traffic, then multiple routes can be used between the EE endpoints (even when multipath routing is not being used). If one of these routes cannot reach the partner EE node, then the EE connection might not activate; if it does activate, then one or more RTP pipes might not activate.
Figure 1. Enterprise Extender with policy-based
routing

The following configuration information applies to Figure 1:
- Policy-based routing is being used with the following definitions:
- A routing rule with the name EEROUTINGRULE1 is defined for EE traffic using EE UDP ports 12003 and 12004 (medium and low traffic priority data). This routing rule is associated with a routing action that points to route table EETABLE1. This route table has a statically defined IP route that uses the OSA1 interface and transmits all data to a next-hop IPv4 address 9.67.1.2 (into IP Cloud1).
- A routing rule with the name EEROUTINGRULE2 is defined for EE traffic that uses EE UDP ports 12000, 12001, and 12002 (LDLC signal; network and high traffic priority data). This routing rule is associated with a routing action that points to route table EETABLE2. This route table has a statically defined IP route using the OSA2 interface and transmits all data to a next-hop IPv4 address 9.67.1.3 (into IP Cloud2).
- Multipath routing is disabled in host HostA.
- Two QDIO OSA interfaces are defined and active in host HostA.
- Two static routes are defined between host HostA and host HostB.
The IP route through IP Cloud1 has a router in the path that has incorrect routing definitions to host HostB. A router in IP Cloud1 with IP address 9.67.1.21 is the router that is unable to route to IP address 9.67.1.6 (HostB).
- The IP route through IP Cloud2 has correct routing definitions to IP address 9.67.1.6 (HostB).
- Two static routes are defined between host HostA and host HostB.