User manual

Release Notes for Catalyst 6500 Series Content Switching Module Software Release 2.2(8)

78-12569-16

Caveats

–

Multiple routes on the CSM to the same destination are supported, but all such routes need to

go through gateways on the same VLAN. Ensure that all routes to any particular destination go

through the same VLAN. For example, the following configuration is invalid because it is

possible for traffic from a remote source to arrive on both VLAN 10 and VLAN 20:

Router(config)# module csm 4

Router(config-module-csm)# vlan 10 client

Router(config-slb-vlan-client)# gateway 1.1.1.1

Router(config-module-csm)# vlan 20 client

Router(config-slb-vlan-client)# gateway 2.2.2.2

To make this configuration valid, delete the gateway command from either VLAN 10 or VLAN

20.

Note NOTE: Do not use the gateway command in more than one VLAN.

–

Traffic into the CSM from an IP address must arrive on the same VLAN that the CSM uses to

send to that IP address. Ensure that all traffic received by the CSM from a specific destination

address arrives on the same VLAN that the CSM uses to reach that destination if the CSM is

configured with a route as follows:

Router(config)# module csm 4

Router(config-module-csm)# vlan 10 client

Router(config-slb-vlan-client)# route 10.0.0.0 255.255.255.0 gateway 1.1.1.1

For load balancing to function properly, all traffic arriving from the 10.0.0.0/24 subnet must

reach the CSM through VLAN 10.

Resolved Caveats in Release 2.2(3)

This section describes caveats that have been resolved in CSM software release 2.2(3).

• CSCdx78285

Under heavy load, it is possible for the CSM to reject incoming connections in the mistaken belief

that it is low on resources. If this issue occurs, then the line entitled Pending event: Re-use too soon

in the output of the show module csm <slot> tech proc 1 command will be increasing rapidly. This

issue exists in CSM releases 1.x, 2.1(1), 2.1(2), 2.1(3), 2.2(1), and 2.2(2a).

Workaround: None.

• CSCdx66816

Active-active collisions are caused by a serverfarm probe failing. In a complex configuration or

under heavy replication traffic, the CSM might exchange its fault tolerance states. This situation

causes temporary blocking of the keep-alive fault tolerance messages. The problem is fixed in

release 2.2(3).

Workaround: For previous software releases in which this issue occurs, increase the failover time

from 3 (the default) to 10 keep-alive messages using the failover command in the ft submode.

• CSCdx66674

When the CSM receives a UDP frame with a 0 checksum on which a NAT operation is to be

performed, the CSM will erroneously overwrite the checksum, causing a UDP frame with an

incorrect checksum to be transmitted. This issue exists in CSM releases 2.1(1), 2.1(2), 2.1(3), 2.2(1),

and 2.2(2a).

Workaround: None.