Re: cam timer tuning

"A router with a stale arp table can't send direct unicast frames to the switch to reach mac addresses no longer in the cam table. The switch would handle these as unknown unicasts and flood them"

Have you passed the ccie r/s written exam? Do you understand and command the theory of the technologies?

I think you may have gone to far without properly mastering the theory.

Joe

From: Naufal Jamal [mailto:naufalccie_at_yahoo.in]
Sent: Friday, December 28, 2012 04:36 PM
To: Marko Milivojevic <markom_at_ipexpert.com>; Joseph L. Brunner
Cc: marco207p_at_gmail.com <marco207p_at_gmail.com>
Subject: Re: cam timer tuning

Understood the unifying timers should definitely work. But if someone asks the theory or the reason behind this fix what could be the explanation?

________________________________
From: Marko Milivojevic <markom_at_ipexpert.com>
To: Joseph L. Brunner <joe_at_affirmedsystems.com>
Cc: "naufalccie_at_yahoo.in" <naufalccie_at_yahoo.in>; "marco207p_at_gmail.com" <marco207p_at_gmail.com>
Sent: Saturday, 29 December 2012 3:00 AM
Subject: Re: cam timer tuning

Truth spoken.

--
Marko Milivojevic - CCIE #18427 (SP R&S)
Senior CCIE Instructor - IPexpert
:: This message was sent from a mobile device. I apologize for errors and brevity. ::
On Dec 28, 2012, at 13:18, "Joseph L. Brunner" <joe_at_affirmedsystems.com<mailto:joe_at_affirmedsystems.com>> wrote:
You're over thinking your network man.
Just unify the cam timer with the arp timer on every interface vlan.
I never have issues and I have a 15,000 node network
From: Naufal Jamal [mailto:naufalccie_at_yahoo.in]
Sent: Friday, December 28, 2012 04:13 PM
To: Marko Milivojevic <markom_at_ipexpert.com<mailto:markom_at_ipexpert.com>>; Joseph L. Brunner; Joe Sanchez <marco207p_at_gmail.com<mailto:marco207p_at_gmail.com>>
Subject: Re: cam timer tuning
http://www.cisco.com/en/US/products/hw/switches/ps700/products_tech_note09186a00801d0808.shtml#cause2
it says,
 In MSFC IOS, there is an optimization that will trigger VLAN interfaces to repopulate their ARP tables when there is a TCN in the respective VLAN. This limits flooding in case of TCNs, as there will be an ARP broadcast and the host MAC address will be relearned as the hosts reply to ARP.
if the ARP table re-populates in an event of TCN, why we need to clear the arp table to fix it?
I know making the cam timer more than the arp timer in 4948 should fix this. but I am not able to justify how it is going to fix it as I think i am missing something on the working of arp in this scenario...
Thanks for your responses so far :)
________________________________
From: Naufal Jamal <naufalccie_at_yahoo.in<mailto:naufalccie_at_yahoo.in>>
To: Marko Milivojevic <markom_at_ipexpert.com<mailto:markom_at_ipexpert.com>>; Joseph L. Brunner <joe_at_affirmedsystems.com<mailto:joe_at_affirmedsystems.com>>; Joe Sanchez <marco207p_at_gmail.com<mailto:marco207p_at_gmail.com>>
Sent: Saturday, 29 December 2012 2:19 AM
Subject: Re: cam timer tuning
adding to this
http://www.cisco.com/en/US/products/hw/switches/ps708/products_tech_note09186a00807347ab.shtml#broadcast
in this URL, i see a note which says
Note: In any Catalyst environment that runs a Hot Standby Router Protocol (HSRP), it is recommended that you ensure the CAM and ARP timers are synchronized
Can anyone please put some light on it ?
________________________________
From: Naufal Jamal <naufalccie_at_yahoo.in<mailto:naufalccie_at_yahoo.in>>
To: Marko Milivojevic <markom_at_ipexpert.com<mailto:markom_at_ipexpert.com>>; Joseph L. Brunner <joe_at_affirmedsystems.com<mailto:joe_at_affirmedsystems.com>>; Joe Sanchez <marco207p_at_gmail.com<mailto:marco207p_at_gmail.com>>
Sent: Saturday, 29 December 2012 2:12 AM
Subject: Re: cam timer tuning
Folks,
(hsrp primary
STP root)
CSW01---------CSW02
     |                  /
     |                /
     |              /
     |            /
      ASW
Lets imagine we are bringing up a new ASW
1) As soon as ASW trunk link is brought up, port in csw01 transitions from blocking to forwarding
2) This initiates STP recalculation. CSW01 sends the TCN BPDU with the TC bit set to all other ASW's connected to it (there are nearly 20 ASW's connected to the CSW's)
3) All the ASW's and CSW's flush out their existing CAM table realizing there is a topology change.
4) All the switches in the L2 domain broadcast arp requests to all the forwarding ports to re-populate the CAM table.
5) For example ASW 20 send ARP for say IP 1.1.1.1, it hits the CSW01 which says the arp mapping 1.1.1.1--->a.a.a.a.
6) it looks for CAM table for a.a.a.a and doesn't find any entry and floods the switchports.
Blogs and organic groups at http://www.ccie.net