Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Community Member

VSS Dual Active dedection

Hi there are several ways to configure

DUAL-ACTIVE-DEDECTION on a VSS-bundle

- (VSL-Link)

- Fast Hello

- Extended PAGP

- IP BFD

How many of them should we implement ? Does it make sense to implement all of them ?

Is there a priority ot tibreaker which of the methods has the last decision about dual-active ?

Thx

Hubert

11 REPLIES
Hall of Fame Super Silver

Re: VSS Dual Active dedection

Hello Hubert,

according to documentation you can use multiple methods at the same time

The virtual switching system supports these three methods for detecting a dual-active scenario:

•Enhanced PAgP-Uses PAgP messaging over the MEC links to communicate between the two chassis through a neighbor switch. Enhanced PAgP is faster than IP BFD, but requires a neighbor switch that supports the PAgP enhancements.

•IP Bidirectional Forwarding Detection (BFD)-Uses BFD messaging over a backup Ethernet connection. IP BFD uses a direct connection between the two chassis and does not require support from a neighbor switch.

•dual-active fast-hello-Uses special hello messages over a backup Ethernet connection. Dual-active fast-hello is faster than IP BFD and does not require support from a neighbor switch. This method is available only in Cisco IOS Release 12.2(33)SXI and later releases,

You can configure all three detection methods to be active at the same time.

see

http://www.cisco.com/en/US/docs/switches/lan/catalyst6500/ios/12.2SX/configuration/guide/vss.html#wp1063718

Hope to help

Giuseppe

Community Member

Re: VSS Dual Active dedection

Thx, I know that I can configure all three methods, but does it make sense ?

And i I configure all three is there a tiebreaker, if e.g. enhanced pagp dedects dual-active and fast hello does not ?

Hubert

Hall of Fame Super Silver

Re: VSS Dual Active dedection

Hello Hubert,

all methods can detect dual-active condition.

I would prefer fast hellos with a dedicated link between the two chassis, then BFD over a dedicated backup link.

I don't like the idea to be dependent on a third switch over data etherchannels with enhanced PAGP.

In this way you have a dedicated link for signalling separated from the forwarding plane.

I would expect the first one to detect the dual active state to win to be enough to move to recovery action.

Hope to help

Giuseppe

Community Member

Re: VSS Dual Active dedection

Anyone know what line cards are supported for IP BFD and dual-active fast-hello?? Is it any line card that is supported on a VSS chassis (any 67XX) or does it need to be specifically 10 GE? I'm thinking not on the latter but I thought I'd ask.

Bronze

Re: VSS Dual Active dedection

We chose to utilize Dual-Active Fast Hello (requires SXI Code) as it's sub-second and doesn't require a configuration on a neighbor.

Running all three, seems like that would be something to shy away from, I'm guessing there would be overhead, etc to worry about. In theory you have a portchannel between the two, running across diversified paths (of course), in addition to another line that can detect if both of the additional lines go down. If you have 10Gb line cards, moving one of the VSL links to a different card is also recommended (We don't have one yet).

For the Dual Active link, we're using a different line card and a Gig Copper SFP. I'm assuming it doesn't matter which card you choose as long as the two can talk to each other. I didn't see anything specific in the configuration guide or Networkers '08 documentation. As long as the chassis can support the card, it's not running VSS, it's just there to check on it.

Hope this helps rate if it does,

JB

Community Member

Re: VSS Dual Active dedection

Thanks man...good to see you on here. I didn't think 10 Gig was required but didn't see anything as well in the doc.

Bronze

Re: VSS Dual Active dedection

I caught you on another thread too. :)Great material on here, great opportunity to learn and share.

Talk to you later,

JB

Hall of Fame Super Silver

Re: VSS Dual Active dedection

Hello Andrew,

a dedicated 10GE link looks like not needed for BFD

see

Dual-Active Detection Using IP BFD

To use the IP BFD detection method, you must provision a >>direct Ethernet >> connection between the two switches. Regular Layer 3 ping will not function correctly on this connection, as both chassis have the same IP address. The virtual switching system instead uses the Bidirectional Forwarding Detection (BFD) protocol.

http://www.cisco.com/en/US/docs/switches/lan/catalyst6500/ios/12.2SX/configuration/guide/vss.html#wp1063892

Hope to help

Giuseppe

Cisco Employee

Re: VSS Dual Active dedection

All dual-active detection protocol are media independent. In another words it can operate on any type of interfaces.

There is no tiebreaker function among this protocols. They all operate independently.

ePAgP and Fast-Hello are rapid dual-active detection protocols that will help in rapid network recovery.

Due to the way BFD function, detection process is slow hence slower network convergence.

Recommendation is to use ePAgP and Fast-Hello combined if possible. If not possible then use multiple trusted PAgP EtherChannel or multiple fast-Hello physical links between both virtual-switch node. You can have up to 4 fast-hello links in an VSS domain...

thanks,

rahul.

Community Member

Re: VSS Dual Active dedection

Hi , I have implemented pagp and fast-hello. I am using a gig link for fast-hello (gi1/6/24 -- gig2/6/24)

When implementing fast-hello the output of the show command is ...

show switch virtual dual-active fast-hello
Fast-hello dual-active detection enabled: Yes

Fast-hello dual-active interfaces:
Port       Local State    Peer Port    Remote State
---------------------------------------------------
Gi1/6/24   Link dn        -            -    

It does not look right does it ????

/Magnus

Cisco Employee

Re: VSS Dual Active dedection

Hello Magnus,

It appears you have configured dual-active fast-hello, however your physical connection between Gi1/6/24 and Gi2/6/24 is down. Once its resolved, the fast-hello adjacency will be formed between both virtual-switch node. You may refer to chapter 2 from following CVD:

http://www.cisco.com/en/US/docs/solutions/Enterprise/Campus/Borderless_Campus_Network_1.0/Borderless_Campus_1.0_Design_Guide.pdf

thanks,

rahul.

814
Views
8
Helpful
11
Replies
CreatePlease to create content