05-01-2015 12:32 AM - edited 05-01-2015 04:38 AM
I have a stack with four members (S1500).
When the primary member is down, the client who is connect to third stack member is lose connection (5-10 ping lost).
It's a normal operation? When a stacking member is down, the clients who connect to other switch in stack lose connection some second?
member-id 0 serial-number "XXX" role primary-capable
member-id 1 serial-number "XXY" role primary-capable
member-id 2 serial-number "XXZ" role line-card
member-id 3 serial-number "XXI" role line-card
add stacking interface stack 0/46
add stacking interface stack 0/47
05-01-2015 06:34 AM
Yes, this is expected behavior. When the PRIMARY member goes down, the SECONDARY will take its place. This process requires resyncronization of protocols suchs as STP and OSPF which will cause a momentary outage.
05-01-2015 01:44 PM
thanks for your reply. When I test linecard member down, i found 1 ping lose, but after the same linecard member is boot up 5-10 ping lose.
In this case why was "long" downtime?
05-01-2015 01:54 PM
Can you give more details on the stack topology (show stacking members and show stacking topology), where you're pinging (client to client within the stack, out an uplink, etc) and which member you are taking down. Depending on the member going down, it can be in the traffic path and so there has to be a re-route through the stack links, or it could be causing a STP event if portfast isn't enabled. There are a number of variables here.
05-02-2015 05:38 AM
Please see the attachments! (network topology, stack informations).
the clients ports set:
The problem is : when the stack member 3 down, 1 ping lose on the clients (it's no problem), and when boot up 5-10 ping lose (it's the problem)! Can I configure something, which reduce this downtime?