Adding new switch to enviroment causes temporary connection drop

[Lisytiro]

We're relative new to cisco switches and are testing two new C9200 switches. At the moment we're trying to setup a simple truck between these two switches, but every time we plugin the cable on the trunk port, one switch drops all icmp-packages for approx. 5 sec. After that the trunk works fine.

our lab contains just these two switches and two laptop with static IPs on both ends for the ping-test.
Modell:C9200-48P
SW Version: 17.06.04
STP: Rapid-PVST
Trunk: Mode on, Encapsulation 801.1q, Status trunking, Native vlan 1
Laptop (Win10)

What could be the cause for one switch to temporary halt all connection for a short time?

[MHM Cisco World]

that normal it STP recover,
immediate when port see BPDU it will go to BLK and check BPDU with one receive from root and 
then it will contious BLK or go to FWD, 
in your case it go to FWD.

that normal I think.

[Bryson F]

just as an example you can do "Switch# show spanning vlan #" i believe it to be the 4th line it will show you the timers on it. I believe the default hello time to be 2 seconds some one please correct me if i am wrong so i don't spread false information. but generally that time shouldn't be an issue, it is normal

[Lisytiro]

thanks for your reply. Im not sure if the BPDU is the problem here. The unexspected drop only happends on connections on the same switch. 

e.g.: if trunk not connected
ping1: PC1 connected to SW1#gi1/0/3 in access mode pings SW1 mgmt-ip -> OK
ping2: PC1 connected to SW1#gi1/0/3 in access mode pings SW2 mgmt-ip -> no reply, cause no physical link (-> OK)
If I now plug in the trunk SW1#gi1/0/1 <-> SW2#gi1/0/1 Ping1 loses 3-5 replys, after that both ping1 and ping2 work (as aspected)

Why should STP move SW1#gi1/0/3 to BLK, when it not receives the BPDU on that port. Shouldn't it only affect the trunk-port (SW1#gi1/0/1), which should not affect ping1?

[MHM Cisco World]

STP not move the g1/0/3 to BLK as I think 
after connect trunk 
ping1: PC1 connected to SW1#gi1/0/3 in access mode pings SW1 mgmt-ip <<<Ping1 loses 3-5 replys, because the CPU is busy in STP recover process 


ping2: PC1 connected to SW1#gi1/0/3 in access mode pings SW2 mgmt-ip <<<this sure will stop because now the SW1 have two path to SW2 and  must elect which one be FWD and which be BLK to prevent Loop.

[bunjiega]

Does gi1/0/3 have portfast enabled? This helps maintain connectivity during the RSTP sync process that would happen when you connect up a new trunk link.