Virtual IP failover not recommended

The appliance handles node failure in WCCP cache farms. If one node goes down, its load is redistributed among the remaining nodes.

In WCCP, you can use multiple routers. Traffic flowing through multiple routers can share the same pool of caches.

In Figure 4, appliances 1, 2, and 3 form a WCCP cache farm.

If the appliance in the WCCP-enabled routing scheme has an ARM bypass rule, the rule causes the appliance to forward particular client requests directly to the origin server, bypassing the appliance. Bypassed requests are unchanged by the ARM; they retain their client source IP addresses. See Adaptive interception bypass‚ on page 126 for details.

In WCCP 2.0, you can exclude certain router interfaces from redirection. The appliance bypass rules can work if you exclude the router interface on which it is connected from using WCCP. To do so, set the router configuration command ip wccp redirect exclude in (refer to Cisco’s WCCP documentation for information about router configuration).

If a WCCP router serves several nodes, as in Figure 4, the router balances their loads. The router sends each node requests aimed at a particular range of IP addresses, so that each node is responsible for caching content residing at particular IP addresses.

The appliance also supports cache affinity. If a node fails and then restarts, the appliance returns the node to its former load distribution. The node’s cache need not be repopulated.

The WCCP cache farm acts as a simple form of distributed cache. A WCCP- enabled network device distributes traffic to individual appliances based on the IP address of the destination Web server. Each node caches objects requested from a particular set of Web servers, which belong to that node’s assigned range of destination IP addresses.

If you are running clustered appliances, Intel recommends that you do not enable virtual IP failover in WCCP environments. The appliance’s WCCP failover mechanism handles node failures and restarts. See Virtual IP failover‚ on

page 146 for details about virtual IP failover.

Using policy-based routing to filter transparency requests

Instead of the WCCP protocol, you can use the policy-routing capabilities of a router to send traffic to the Intel NetStructure Cache Appliance. WCCP or an L4 switch is generally preferable to policy-based routing because it has a performance impact on the router and does not support load balancing or heartbeat messaging.

124Intel NetStructure Cache Appliance Administrator’s Guide

Page 136
Image 136
Intel 1520 manual Using policy-based routing to filter transparency requests