6509 virtual switching system

I just wanted to know if anyone in these forums has any experience with this technology. Our sup's are capable of this at work, and a consultant is currently exploring options with it for us (we have dual 6509's). Has anyone ever had any issues with it? The configuration looks like it isnt too bad, and having 1 management console would be outstanding. Truthfully, this is the first I have ever heard of the technology, and am trying to read white paper after white paper about it because it is facinating, but wanted some insight from the great minds at techexams.net:).

Find more posts tagged with

Free for TechExams community: Cybersecurity salary guide

Compare cert salaries and plan your next career move

Button

Comments

adam-b

I had not even heard of VSS until you posted this...awesome! I was immediately curious about it since we have two 6509 core switches each housing a FWSM in active/standby routed mode. It looks like we can use VSS without interfering with the FWSM's mode of operation which was my initial concern. I am going to explore this more, thanks for bringing this technology to discussion. Hopefully somebody hear has or is using it and can provide additional insight.

joshgibson82

jason_lunde wrote: »

Truthfully, this is the first I have ever heard of the technology, and am trying to read white paper after white paper about it because it is facinating, but wanted some insight from the great minds at techexams.net:).

If you have a CCO login, there should be loads of configuration guides/examples on cisco.com for this. Our company is moving to it, but have not deployed it yet to my knowledge.

dtlokee

I have installed several of them and there have been different issues with each one. Since you already have the switches installed when you convert them to VSS you lose your configuration.

I guess the question I have is why do you want to go to VSS? What is the business requirement that VSS will solve? If the answer is to only have a single management interface for both switches you may want to reconsider due to the amount of effort that is required to do the conversion.

Cyanic

We are looking at it for cleaner redundancy over HSRP, but we have a a while to look into it. We need to upgrade our Sup 720s first. If we do go this route, we will be implementing it at 7 locations on our network.

jason_lunde

Right, I think we are really looking to eliminate HSRP, and to take full advantage of the SUP's. We also have a while to look into this, but thanks for bringing up that it totally wipes the configs DT (this is why I enjoy these forums so much). I obviously have some more reading to do, and am anxiously awaiting our consultants report. I am also finding that basically our distribution switches will also be able to effectively have 1 run to each of the physical cores, but since the cores will be seen as 1 logical device these separate runs can be etherchanneled together. Is this correct?

astorrs

jason_lunde wrote: »

Right, I think we are really looking to eliminate HSRP, and to take full advantage of the SUP's. We also have a while to look into this, but thanks for bringing up that it totally wipes the configs DT (this is why I enjoy these forums so much). I obviously have some more reading to do, and am anxiously awaiting our consultants report. I am also finding that basically our distribution switches will also be able to effectively have 1 run to each of the physical cores, but since the cores will be seen as 1 logical device these separate runs can be etherchanneled together. Is this correct?

Yes, the VSS operates as one big switch, thereby allowing you to have trunks to separate chassis (for redundancy like you probably do now) yet unlike your current setup both will be active and passing data. The best of both worlds.

dtlokee

astorrs wrote: »

Yes, the VSS operates as one big switch, thereby allowing you to have trunks to separate chassis (for redundancy like you probably do now) yet unlike your current setup both will be active and passing data. The best of both worlds.

You always could trunk to both chassis in the distribution layer or core layer and you could load balance across them by pruning the VLANs on the trunks or by designating the 2 distribution/core layer switches as root for different VLANS. What is different now is that you can create an etherchannel group that spans across both switches in the VSS group. Neither form of load balancing is perfect, VLAN pruning is labor intensive and will need retuning to keep it "balanced" as traffic changes, Etherchannel will hash the header and use that to determine what link to send it over, which can result in unequal load balancing.

For VSS to be most effective you must dual home everything into both switches.

astorrs

dtlokee wrote: »

You always could trunk to both chassis in the distribution layer or core layer and you could load balance across them by pruning the VLANs on the trunks or by designating the 2 distribution/core layer switches as root for different VLANS. What is different now is that you can create an etherchannel group that spans across both switches in the VSS group. Neither form of load balancing is perfect, VLAN pruning is labor intensive and will need retuning to keep it "balanced" as traffic changes, Etherchannel will hash the header and use that to determine what link to send it over, which can result in unequal load balancing.

For VSS to be most effective you must dual home everything into both switches.

That's what I meant, VLAN pruning is a lot of work (but it's what most of my clients have ended up doing) and dual homing everything into both switches gives you both simplicity (comparatively) and redundancy.

APA

I remember VSS being brought up as the 'next-big thing' in the Cisco world at a Tech Solutions briefing a while back.

Good to see people are latching onto it.

As yet I haven't had to implement it yet, more of a nice to have than need to have especially when you've had kit implemented already as dt mentioned

apd123

Implementing it in a new build was not too much of an issue. This was a fairly easy technology to go from I have no idea what that is to implementing it. If I had to do it during an outage window on a built config with hardware and IOS upgrade to support VSS it might not have been as fun.