[c-nsp] carving TCAM on Engine 3 linecards due to exhausted IPv6 TCAM regions
Per Carlson
pelle at hemmop.com
Fri Feb 18 03:52:48 EST 2011
Hi.
On Fri, Feb 18, 2011 at 07:46, Jan Sandmaier <sandmaier at schlund.net> wrote:
> Hi Pelle,
>
> thanks for your good answer. I simply can't configure this. What code
> are you running? I'm running an quite old code. It seems I have to do an
> upgrade first.
We are running SY9b.
> bb-b.fra3.fra.de(config)#hw-module slot 1 tcam carve ?
> RX_144b_REGION RX 144b Dynamic Region
> RX_288b_REGION RX 288b Dynamic Region
> RX_72b_REGION RX 72b Dynamic Region
> RX_BOT_144b_REGION RX Bottom 144b Dynamic Region
> RX_BOT_288b_REGION RX Bottom 288b Dynamic Region
> RX_BOT_72b_REGION RX Bottom 72b Dynamic Region
> RX_IPV6_MCAST_288b_REGION RX IPv6 Mcast 288b Region
> RX_IPv6_144b_REGION RX IPv6 144b Region
> RX_TOP_144b_REGION RX TOP 144b Fixed Region
> RX_TOP_288b_REGION RX TOP 288b Fixed Region
> RX_TOP_72b_REGION RX TOP 72b Fixed Region
> RX_TOP_NF_REGION RX TOP NetFlow Region
> TX_144b_REGION TX 144b Dynamic Region
> TX_288b_REGION TX 288b Dynamic Region
> TX_72b_REGION TX 72b Dynamic Region
> TX_BOT_144b_REGION TX Bottom 144b Dynamic Region
> TX_BOT_288b_REGION TX Bottom 288b Dynamic Region
> TX_BOT_72b_REGION TX Bottom 72b Dynamic Region
> TX_IPv6_144b_REGION TX IPv6 144b Region
> TX_TOP_144b_REGION TX TOP 144b Fixed Region
> TX_TOP_288b_REGION TX TOP 288b Fixed Region
> TX_TOP_72b_REGION TX TOP 72b Fixed Region
> TX_TOP_NF_REGION TX TOP Netflow Region
When looking into it more closely, we do neither have any RX_IPV6
region to configure:
router(config)#hw-module slot 0 tcam carve ? | filter
RX_IPV6_MCAST_288b_REGION RX IPv6 Mcast 288b Region
RX_IPv6_144b_REGION RX IPv6 144b Region
RX_TOP_NF_REGION RX TOP NetFlow Region
TX_IPv6_144b_REGION TX IPv6 144b Region
TX_TOP_NF_REGION TX TOP Netflow Region
router(config)#
It very looks like someone have pasted the wrong commands into the Bug-Id.
*I* would assume it'e the RX_IPv6_144b_REGION that should be used
together with a 128 bit prefix, but a confirmation of a Cisco employee
on this list would be nice (poking Rodney or Oliver ;-) ).
--
Pelle
RFC1925, truth 11:
Every old idea will be proposed again with a different name and
a different presentation, regardless of whether it works.
More information about the cisco-nsp
mailing list