[c-nsp] XENPAK 10GB-LW (WAN PHY)

Richard A Steenbergen ras at e-gerbil.net
Fri May 26 05:13:00 EDT 2006


On Fri, May 26, 2006 at 10:02:09AM +0300, Emanuel Popa wrote:
> Hi,
> 
> Is there anybody using 10GE WAN PHY with WS-X6704-10GE and
> XENPAK-10GB-LR in a Cisco 7600 platform?
> 
> We are using 2 pieces to connect 2 routers via a carrier DWDM network.
> The trouble is that the interface is flapping very often and the
> STM-64 line test is error free. There is no support on the Ethernet
> interface for SDH errors like LOS, LOF, AIS, RDI etc. We can't figure
> out what the problem might be.

I've seen that behavior a lot with the WAN PHY XENPAKs in 6704s. It seems 
that the non-Cisco optics (i.e. Optillion produced optics used by folks 
like Foundry etc) will never come up, they'll just flap like that forever. 
Even with the offical Cisco XENPAKs, any error on the interface or 
receiving a signal that is just outside of spec seems to trigger 
uncontrolled flapping until the port is errdisabled. The lack of SONET 
diagnostic capabilities in any form was disturbing as well.

The weird thing is that the SP will continue to see the flaping even with 
the interface shutdown. Within a couple of hours this triggers a state 
where the box thinks the transceiver is being removed and reinserted 
(until sometimes it wedges in a completely missing state :P). Of course if 
you have any non-Cisco optics, every time this happens to generates the 
nag messages for them, filling your logs.

> Besides the problems above, there are two more issues:
> 
> 1. the thresholds for miliamperes - current for this XENPAK are not
> configured properly in this IOS and they can not be configured via
> CLI:
> 
>                            High Alarm  High Warn  Low Warn   Low Alarm
>         Current            Threshold   Threshold  Threshold  Threshold
> Port     (milliamperes)     (mA)        (mA)       (mA)       (mA)
> -------  -----------------  ----------  ---------  ---------  ---------
> Te11/4    28.7         ++      2.5         2.5         2.5        2.5

The DOM data for the LW's had a tendancy to be very very wrong the last 
time I tested it (all of this is under SXF3). I was able to get it into a 
state of misreporting pretty much everything with minimal playing around, 
including reporting signal when a fiber wasn't connected and even when the 
transceiver wasn't plugged in at all. :)

-- 
Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)


More information about the cisco-nsp mailing list