IGX 84xx clock problem over satellite, but not in lab.

[mhaaijer]

Hello,

Between 3 IGX 84xx node network (triangle) there seems to be a problem with the clock on one switch. The network is very unstable and is more busy with itself than passing traffic. There are lots of failures on the trunk as a result. Is this a SCM problem or..... All trunks go via satellite.

Thanks for helping!

Mike

[smalkeric]

I found this on Cisco's site.. might be useful to you: http://www.cisco.com/warp/public/74/topissues/wan_sw/clocking.html

[Michael Simon]

Mike,

Passing clocks across a satellite link is typically challenging. The reasons are that though the satellites appear to be stable platforms, they move constantly relative to a point on the earth. As a result, there is a degradation of the clocking information due to Doppler shift.

To better answer your exact situation I need some more information:

1. What is the clock reference for your three nodes? each node? (dspcurclk and dspclksrcs)

2. Are you attempting to pass clocks across the satellites (i.e. pass sync = Y on cnftrk)?

3. Are you using your own v-sat terminals or are these links provided thru a commercial telco? If your own - what satellite modem and does it have built-in buffers? Which telco?

........Mike

[mhaaijer]

Hi Mike,

Thanks for the info. The clocks are set to be correct, so that should not be that problem. The info on the equipment on the trunks is as follows:

These are our own ComTech EFdata modems type SDM 300A with the options Reed Solomon and Drop & Insert. The channel is configured as E1-CCS with HDB3 G703 unbalanced.

Mike