cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
905
Views
5
Helpful
7
Replies

HSRP and stack together?

abtt-39
Level 1
Level 1

Hello,

Is it possible to use hsrp and stack at the same time?

We have a core switch which is a stack of 4 switches

We would like to add a second stack for redundancy.
2 stacks in 2 different buildings with 9300 switches.With hsrp in between ( active / standby).

how long does failover take in a stack, by default? for example if the master fail?

And same question with HSRP, i'v read that If a hello has not been received for 10 seconds by the standby then it becomes active?

But my question is whether these 2 redundancy methods can coexist together?

7 Replies 7

Hi

 HSRP failover time is 10 seconds.  Inside a stack, the failover must be instantaneously.

Stack is meant to be Physical redundancy and HSRP is meant to be Layer 3 redundancy. They can coexist with no problem. In your scenario, you will end up with two stacks of 9300 and each one of them will have its own Ip address. 

  Both stacks must share the same Vlan and the Interface vlan must be on the same subnet. You can save the IP address final .3, for example, for one stack, the final .2 for the other stack and the final .1 to be the standby for the interface vlan and the gateway for your clients. the final .1, on this case, will be the VIP IP for the HSRP.

If stack one fail the active will the the stack 2 and vice-versa

Thank you Flavio

I have another question.
When I change, for example, a configuration, add something on stack 1 (active),
do I have to do it systematically on stack 2 (stand by)? or is there some kind of configuration replication?

 

Hello,

Is it possible to use hsrp and stack at the same time?

We have a core switch which is a stack of 4 switches

We would like to add a second stack for redundancy.
2 stacks in 2 different buildings with 9300 switches.With hsrp in between ( active / standby).

how long does failover take in a stack, by default? for example if the master fail?

And same question with HSRP, i'v read that If a hello has not been received for 10 seconds by the standby then it becomes active?

But my question is whether these 2 redundancy methods can coexist together?

Two stack switches are two independent devices, they will not share config or sync config. think about them as two individual devices

 
 
stack and HSRP can work together no problem 

Joseph W. Doherty
Hall of Fame
Hall of Fame

In theory, you should be able to do both.

In practice, likely depends on the switch's IOS features.  For example, I recall HSRP wasn't supported on the 3750 series (or perhaps not supported on the early IOS versions), but it may be supported on your 9300.  One great way to tell, is check the Cisco documentation for your particular platform and IOS release.

"how long does failover take in a stack, by default? for example if the master fail?"

NB: my off-the-top-of-my-head info, is very dated (i.e. 3750 stacks), but some issues may still remain some may not.

So, you're asking about an actual stack master failure?

Recovery time, may depend on whether we're discussing L2 or L3 transiting the switch, and recovery for dynamic routing protocols.  (Overall, I would expect the latest Cisco Catalyst switches, to be "faster" than earlier models.)

One issue used to be, whether (w/o a FHRP virtual IP and virtual MAC) whether gateway MACs changes with the master recovery (they did [again on 3750s] by default, but you could configure replacement master to use prior gateway MAC).

L3 dynamic routing might see the L3 "router" drop, unless you configured, correctly, NSF, on your stack and adjacent "routers" supported it.

At worst (again 3750 stack), a master hit might take the whole stack off-line for close to 30 seconds (or so I recall), at best, using all the (optional) fast failover features, there's no hit to traffic.  (NSF - non stop forwarding)

"And same question with HSRP, i'v read that If a hello has not been received for 10 seconds by the standby then it becomes active?"

Probably true using default settings, but HSRP timers can be reduced, and with HSRP v2 and/or HSRP with BFD support, you can, I believe, get into subsecond failure detection.

BTW, another new feature in the Catalyst 9Ks, is StackWise Virtual, but don't know if you can use two stacks and/or whether your 9300 supports this feature.

BTW, since you've mention different buildings, a "classical" HSRP design, VLANs per building, might be just fine, but if your VLANs' usage spans the buildings, they you might want to consider host to particular stack mapping, to avoid interbuilding latency.  HSRP v2 support multiple primary gateways, across devices.

If using both buildings' stacks, regardless of host locations isn't a consideration, you might also see if your 9300 stacks supports GLBP.  (Maybe it was GLBP that 3750 stacks didn't support - again, so long ago, unsure.)

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: