Solved: Sup720 SSO redundancy question

poirot1967 · ‎03-09-2008

First off, this could just be my misunderstanding of SSO and redundancy, and I may be trying to fix something that is not broken. I have looked thru the documentation, and searched the forums, but am still a little confused.

We replaced an aging Cat 4k with a 6506 with redundant Sup720's in our data center. Everything seems to be running perfectly, but there is an odd (to me) thing going on. The Sup720 in slot 5 is listed as the active sup, but when looking at the counters, the sup in slot 6 seems to be doing all the heavy lifting. It was my understanding that the active sup would be handling all packet transfers, while the standby was listening to the "heartbeat" waiting to jump in.

I have attached a pdf file the counters, and redundancy info. If I have noting to worry about, great. It just seems odd to me.

Thanks

Poirot

glen.grant · ‎03-09-2008

There isn't a problem I don't believe , slot 6 is standby but even though slot 6 is the standby supervisor the actual uplink ports on both supervisors are active and usuable . If you look at your config you will probably see that slot 6 link configured and it just happens to handling more traffic that the slot 5 uplink, nothing inherently wrong with that.

View solution in original post

glen.grant · ‎03-09-2008

There isn't a problem I don't believe , slot 6 is standby but even though slot 6 is the standby supervisor the actual uplink ports on both supervisors are active and usuable . If you look at your config you will probably see that slot 6 link configured and it just happens to handling more traffic that the slot 5 uplink, nothing inherently wrong with that.

poirot1967 · ‎03-09-2008

Thanks for the reply. I just wanted to be sure I didn't screw something up in the config. It was a nightmare just getting the upgrade plan approved as it was. I read in the documentation that both interfaces would be active, but I didn't think that the standby uplink would be THAT active. Thanks again.

Poirot

poirot1967 · ‎03-12-2008

Just as a follow up.... After a few days of monitoring, I still kinda feel things are a little strange. Looking at the stats in 30 min intervals, the active sup has 0% utilization on its uplink, with a peak TX of 9.9Kps, and a peak RX of 69bps. The hot spare by comparison has 7% RX and 13% TX, with a peak of 250Mbps and 420Mbps respectively.

It looks like the active sup is using the hot spare's uplink instead of its own. If that is the case then everything makes send to me, otherwise, I am stumped. What is your take on it.

Thanks

Poirot

clark.adam.p · ‎03-12-2008

As others have said, the ports are independant of the supervisors. Of the ports are trunks or members of the same VLAN then its completely possible that 6/1 has a lower STP cost than 5/1.

This is completely dependent on your topology which we do not know intimately.

If your so inclined, you could swap positions of the supervisors.

I would imagine that this is something completely benign.

Adam

poirot1967 · ‎03-13-2008

*Ding* and the light goes on....

So the ports are independent of the sup engines....now it all makes sense. I can sleep now...

Thanks

Poirot