Dear All,
I have a scenario where my 2 4500x in VSS are connected via 2 FO links to a Cat 6500.
The 6500 is only one box and it has 2 SUP2T.
The connection has been configured as a l2 PO and inside the PO , 3 VLAN are propagated (vlan 100, 31 and 32).
On the 6500 there are also the SVI while the on the other side the 4500x is just a l2 and the L3 in configured on a firewall.
This set up is up since more than 1 year without any issue... Till some days ago...
we started experiencing packet loss, so as some previous experiences i isoleted the problem that seems to be between the 4500x member 1 and the 6500.
I removed one member from the PO and everything starts working again.
Today... even if there are not any error or elements to point to the physical components i have replaced the sfp+ and the FO from the 4500x member 1 and the 6500 and the problem got even worse.
I don't know why.... but even if both members of the PO where up, the SVI on the 6500 starts flapping:
%LINK-3-UPDOWN: Interface TenGigabitEthernet1/5, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface TenGigabitEthernet1/5, changed state to up
%LINK-3-UPDOWN: Interface Vlan31, changed state to down
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan31, changed state to down
%LINK-3-UPDOWN: Interface Vlan100, changed state to down
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan100, changed state to down
%LINK-3-UPDOWN: Interface Vlan100, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan100, changed state to up
%LINK-3-UPDOWN: Interface Vlan31, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan31, changed state to up
%LINK-3-UPDOWN: Interface Vlan31, changed state to down
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan31, changed state to down
%LINK-3-UPDOWN: Interface Vlan31, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan31, changed state to up
%LINK-3-UPDOWN: Interface Vlan32, changed state to down
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan32, changed state to down
%LINK-3-UPDOWN: Interface Vlan32, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan32, changed state to up
%LINK-3-UPDOWN: Interface Vlan31, changed state to down
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan31, changed state to down
%LINK-3-UPDOWN: Interface Vlan31, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan31, changed state to up
%LINK-3-UPDOWN: Interface Vlan32, changed state to down
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan32, changed state to down
%LINK-3-UPDOWN: Interface Vlan32, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan32, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface TenGigabitEthernet1/5, changed state to down
%LINK-3-UPDOWN: Interface TenGigabitEthernet1/5, changed state to down
the flap was waiting rstp timing...
As far as i know, to get a SVI down there are just few reason:
VLAN exists and is in active status on the switch VLAN database.
VLAN interface exists on the router and is not administratively down.
At least one L2 (access port or trunk) port exists and has a link up on this VLAN. The latest implementation of the autostate feature allows synchronization to Spanning-Tree Protocol (STP) port status
But as you can see in the log... the PO was all the time up... no STP recalculation .....
Anyone experienced something similar?
Daniele
Ps
Attached a small net diagram
