Cat6509 Redundancy 2

[Microbyte]

Hi,

This is the network,

6509a
/ |
A - US |
\ |
6509b

A - computer node
US - unmanaged switch
6509a - Cat6509
6509b - Cat6509

Two Cat6509 are linked via layer-2 trunking.

What is the best method of configuring 6509 to support redundancy?

I appreciate all your help in advance!

Thanks,





Microbyte

 

[sirlollypop]

Manage the switch

But go Layer 3 with the 6509's and run HSRP from the 6500 down to the access switch much quicker convergence than Spamming tree

 

[vipergg]

It will work either at layer 3 or layer 2 . If you are worried about convergence time use Rapid spanning tree , failover is like 1-2 seconds for spanning tree . You slow spot would be the hsrp failover time , you can mess with the timers just make sure they are the same all the way around .

 

[globalchicken]

You also could consider uplink fast. This is also at L2 but failover is instant.

 

[Microbyte]

Thanks to all!

I've decided to go with layer-3 method because I was afraid of spamming tree causing any issue with having two gateways in meshed configuration.

first 6500's configuration,

interface FastEthernet4/1
Ip address 192.1.1.2 255.255.255.0
No ip redirects
Ip pim dense-mode
Duplex full
Standby 100 ip 192.1.1.1
Standby 100 timers msec 100 msec 300
Standby 100 priority 200
Standby 100 preempt


Second 6500's configuration,

interface FastEthernet4/1
Ip address 192.1.1.3 255.255.255.0
No ip redirects
Ip pim dnese-mode
Duplex full
Standby 100 ip 192.1.l.1
Standby 100 timers msec 100 msec 300


after all done, i shutdown the first 6500's port 4/1.
and it shutdown the communication from A node to the switch.
After fiddling around, found
eigrp 1
network 192.1.1.0 255.255.255.0 was missing.

That took care of the failover issue.

but now i am faced with another problem. Not a serious one, but the switch is displaying status change message every 4 seconds going back-n-forth between active and standby

Standby-4-statechange: fastetherenet4/1 group 100 state standby -> active
Standby-4-statechange: fastetherenet4/1 group 100 state active -> standby

Does anybody have any idea why switch is having this issue?
Thanks!


Microbyte

 

[sirlollypop]

can you post a bit of a deubug standby or debug standby errors but I have never used timers that close together. Are you seeing any other errors on either of the two Fast etherents?

 

[vipergg]

Those are awful short hsrp timers , if you get any kind of marginal port connections between switches you are going to get flapping hsrp state changes .

 

[voltron1011]

I agree with Viper...
It looks like your timers are too short, which is probably causing your flapping issues.

 

[lerdalt]

adjusting the HSRP timers can be risky. If you have the ports available, you might want to setup a port-channel between the 2 6509's.

 

[Microbyte]

viper and voltron,
What do you recommend the hsrp timer to be?

Don't want to lose any data as well as avoid any flapping.
If I can't have both, what would be the compromised hsrp timer value?

Sirlollypop,
I didn't do any other debug command. Above command is temporarly removed because of CPU utilization to jump 5 folds.

lerdalt,
Also I do not have choice of creating portchannel between the two.

Thanks all!

Microbyte

 

[lerdalt]

I'd suggest putting the timers back to defaults first and see what happens.

Also, personally, I like setting both priority numbers to something other than the default....just so you can see what it really is in the config. Nothing major, just preference.

 

[vipergg]

If you have that 6509 setup with a crosslink across it allowing all vlans across that trunk and use rapid spanning tree , hsrp should very rarely ever go down . We use timers 5 20 , which has always worked well for us , though not losing any data at all is pretty hard .