cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
750
Views
0
Helpful
4
Replies

6500 VSS Dual Active Situation with 720 quad sups

jbnair
Level 1
Level 1

Hello Guys,

I have the dual active situation, could you tell me how you are running the dual active scenarios and which is better than the other 2 options (i mean BFD, PAGP etc)

Let me add my scenario..

2 x 6509 Switches

Quad 720 Sups (2 each per chassis)

The sups are cross linked with 10gbps LR modules.  The G1/5/3 and G2/5/3 Copper ports in the Active and Standby Sups are connected and configured with dual active BFD configuration.

Now, when I shutdown the Active chassis, the standby is coming up as active -- no prorblem. Then when I swiched on the OLD-Active Chassis back, it is coming up, but now  -- the dual active scenario happens as it shows both chassis runs with Active (green) LED. On the console, it shows the message that the system had found dual active situation with the help of BFD and one of the chassis goes with Recovery mode and we will not able to do anything on the recovery chassis console (the OLD standby chassis shown the Recovery mode).

The point is the recovery mode shows more than 30 minutes and nothing happens. I searched for the min/max dual active recovery time, but documents says only about the dual active detection time such as BFD will detect it in 20-25 seconds.

The end result is that VSS runs as two active chassis and i am sitting wih that without doing other job.. 2 x IDSM, 2 x 5585 are waiting for me to give further troubles, so i would highly appreciate your comments about how you achieved successful dual active recoveries.

Finally, one person has pointed that when he runs a BFD or PaGP dual active system throguh a line card or via MEC with edge switches, everything is working as one is going to standby without any hesitation.

I am using 12.2(33) SXI4 IOS.

Regards

Jay

4 Replies 4

Wilson Bonilla
Level 3
Level 3

Hello Sir

In regards to the first question, I'm running Pagp

And this is how to run BFD

- They are different desing options, don't say any is better than the other.

- A dual-active is triggered in miliseconds, BFP can trigger a switch to go to recovery mode in 300msecs. (It's incredible fast)

In order to help you determine why it is not going into recovery mode please post the following outputs:

Show switch virtual dual-active summary

Show switch virtual dual-active bfp

Show switch virtual dual-active pagp.

Show etherchannel summary

Show log

Please post all useful post.

Regards .

Wilson B

Hi Wilson,

Thank you for the reply and details about the queries.

I have reworked and everything is running smooth now.

The following steps have been taken.

1. Installed the VSS.

2. Created a MEC via a 3750 switch stack.

3. Configured Enhanced PAgP for detecting dual active scnerios, used MEC via 3750 stack.

4. Tested shut down of each chassis. The switchover was perfectly worked as dual detection is really happening via the MEC.

5. I got advised by another person that instead of using the fiber modules on Sups, try it out with 10 Gig Line card modules and that also would work. Anyway my line card is dedicated for connecting some edge stacks, so i decided to stay with this plan.

Cheers

Jay

Hi.

Glad to hear that.

I would like to share with you the following information just for future reference:

Note: If you configure ePagp in a cross-stack etherchannel it will not work since it's not a supported configuration.

Table 4-12     Cisco IOS Version Support for Enhanced PAgP 

Platform
Software
Comments

Cisco Catalyst 6500

12.2(33)SXH

Sup720 and Sup32

Cisco Catalyst 45xx and Cisco Catalyst 49xx

12.2(44)SG

Cisco Catalyst 29xx, Cisco Catalyst 35xx and Cisco Catalyst 37xx

12.2(46)SE

Cisco Catalyst 37xx stack no support, see the text that follows.

Cisco Catalyst 37xx Stack

Not Supported

Cross-stack EtherChannel only supports LACP

http://www.cisco.com/en/US/docs/solutions/Enterprise/Campus/VSS30dg/VSS-dg_ch4.html

Regards.

Wilson B.

Hi Wilson,

Thanks for the reply. I used SXI4 IOS on 6500 and used etherchannel ON mode on the MEC link. Then configured enhanced PAgP pointed this etherchannel and now everything shows cool. I restarted both chassis 2-3 times and everything looks ok.

One thing I noted that when we shutdown the active chassis, the other chassis will immediately come up, but the second Sup is taking the control making the first sup as standby (i am saying about the scenatio where i am having dual sups per chassis). And when we are in the second Sup, it will not show the VSS status as the switch virtual commands are unavailable. And then once we switched on the old Active chassis back, it will be coming as standby, visually we can see this showing the orange LED on the chassis. This condition stays there unless and until we move to the next shutdown and then when you shutdown the current active chassis which works with the second Sup, then things will be back normally, as the first active chassis will take over with its first sup and start showing the switch virtual status.

SXI4 onwards supports dual Sups per chassis and my customer is interested in latest 15.x IOS on 6500, but i am not planning it now until I get some confirmation from forums that it is ok to go ahead.

Regards

Jay

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: