VSL down - Active Role

Answered Question
Aug 13th, 2010

I am testing a few possibilities about Failover with SUP720-VSS

this is my config

switch virtual domain 100
switch mode virtual
switch 1 priority 254



VSL betwenn both 10G Portsof the SUPs as Portchannel


interface Port-channel1
no switchport
no ip address
switch virtual link 1
mls qos trust cos
no mls qos channel-consistency
!
interface Port-channel2
no switchport
no ip address
switch virtual link 2
mls qos trust cos
no mls qos channel-consistency


and one dual-active fast-hello Port


interface GigabitEthernet1/7/8
no switchport
no ip address
dual-active fast-hello


interface GigabitEthernet2/7/8
no switchport
no ip address
dual-active fast-hello

All is working fine.

Now I cut the connections of VSL


Because of dual-active fast-hello detection:

local Switch 1 is going in shutdown

local Switch 2 is getting active

Why the Switch 2 is getting active and not Switch1?

Is there a parameter where I can specify this?

The command "switch 1 priority 254 " seems to be not in conjunction with this.#

I reconnect the VSL connection

On Switch1

Switch  Switch Status  Preempt    Priority  Role     Session ID
        Number         Oper(Conf) Oper(Conf)         Local  Remote
------------------------------------------------------------------
LOCAL    1     UP      FALSE(N )   254(254)  ACTIVE   0      0
REMOTE   2     UP      FALSE(N )   100(100)  ACTIVE   4974   6188


In dual-active recovery mode: Yes
  Triggered by: Fast-hello detection
  Triggered on interface: Gi1/7/8

VSS(recovery-mode)#show switch virtual redundancy mismatch

No Config Mismatch between Active and Standby switches


On Switch2


VSS#show switch virtual role

Switch  Switch Status  Preempt    Priority  Role     Session ID
        Number         Oper(Conf) Oper(Conf)         Local  Remote
------------------------------------------------------------------
LOCAL    2     UP      FALSE(N )   100(100)  ACTIVE   0      0
REMOTE   1     UP      FALSE(N )   254(254)  ACTIVE   6188   4974


In dual-active recovery mode: No


VSS#show switch virtual redundancy mismatch

No Config Mismatch between Active and Standby switches

Why is there no mechanism which recreate a normal functional VSS?

I want to do no interaction in such case.

I have this problem too.
1 vote
Correct Answer by jorge.calvo about 6 years 3 months ago

Hello,

Let me try to answer your questions:

Now I cut the connections of VSL


Because of dual-active fast-hello detection:

local Switch 1 is going in shutdown

local Switch 2 is getting active

Why the Switch 2 is getting active and not Switch1?

This occurrs because during normal VSS operation, prior to the dual-active condition, Switch1 was the Active switch and Switch2 was the Hot-Standby.

When dual-active condition is detected via Fast-Hello mechanism the old-active switch disables all the local interfaces except the loopback to avoid the instability caused by dual-active situation. It also removes routing and STP instances. After that it remains out of the network.

Is there a parameter where I can specify this?

This situation is always the same. The old-active switch will be shutdown and the old-Hot-Standby switch will become active.

I reconnect the VSL connection

Why is there no mechanism which recreate a normal functional VSS?

I want to do no interaction in such case.

Once VSL is restored, Switch1 reloads and comes up in standby mode. Then, RRP will determine that the old-active switch Switch1 must be the Hot-Standby switch now. RRP considers there is no need to change the role of the new-active switch (Switch2) and cause more data loss.

The way the Active and Hot-Standby roles are elected depends on the order the switches are initialized. If they boot up at the same time, the switch with lower switch ID will be the Active switch.

If one of the switches is desired to be the Active one under all conditions, preemption must be configured on it and the desired Active switch must have the highest switch ID.

However, enabling preemption is very dangerous for the network stability because it forces multiple reboots of the VSS member.

Hope this helps.

Correct Answer by Reza Sharifi about 6 years 3 months ago

Marcus,

It is not recommended to use preemption in the data center; typically, the VSS member switches uses identical line card configurations and reside in a dual-homed environment, where favoring one system over another is irrelevant.

HTH

Reza

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (2 ratings)
Loading.
Correct Answer
Reza Sharifi Fri, 08/13/2010 - 06:41

Marcus,

It is not recommended to use preemption in the data center; typically, the VSS member switches uses identical line card configurations and reside in a dual-homed environment, where favoring one system over another is irrelevant.

HTH

Reza

Correct Answer
jorge.calvo Fri, 08/13/2010 - 06:48

Hello,

Let me try to answer your questions:

Now I cut the connections of VSL


Because of dual-active fast-hello detection:

local Switch 1 is going in shutdown

local Switch 2 is getting active

Why the Switch 2 is getting active and not Switch1?

This occurrs because during normal VSS operation, prior to the dual-active condition, Switch1 was the Active switch and Switch2 was the Hot-Standby.

When dual-active condition is detected via Fast-Hello mechanism the old-active switch disables all the local interfaces except the loopback to avoid the instability caused by dual-active situation. It also removes routing and STP instances. After that it remains out of the network.

Is there a parameter where I can specify this?

This situation is always the same. The old-active switch will be shutdown and the old-Hot-Standby switch will become active.

I reconnect the VSL connection

Why is there no mechanism which recreate a normal functional VSS?

I want to do no interaction in such case.

Once VSL is restored, Switch1 reloads and comes up in standby mode. Then, RRP will determine that the old-active switch Switch1 must be the Hot-Standby switch now. RRP considers there is no need to change the role of the new-active switch (Switch2) and cause more data loss.

The way the Active and Hot-Standby roles are elected depends on the order the switches are initialized. If they boot up at the same time, the switch with lower switch ID will be the Active switch.

If one of the switches is desired to be the Active one under all conditions, preemption must be configured on it and the desired Active switch must have the highest switch ID.

However, enabling preemption is very dangerous for the network stability because it forces multiple reboots of the VSS member.

Hope this helps.

Marcus Hunold Tue, 08/17/2010 - 06:49

Because of using the SXI IOS Version there is no problem with preemption because it isn't available...

Thank you for your answers - all of it I could reproduce :-)

Actions

This Discussion