11-27-2018 06:40 AM - edited 03-08-2019 04:41 PM
Hello,
I have a weird issue that when one of my Port Channel comes UP on my 3850, the entire switch freezes for a few seconds.
Basically, I have a two switch stack of WS-C3850-48U and six IE-2000-4TS-G-B connected via Etherchannel on the 3850. All IE-2000-4TS-G-B are configured exactly the same and all Port Channel on the 3850 are identical. For some reason, when I bring the port channel of one of the IE-2000 down and bring it back up, the entire 3850 freezes for a few seconds as the port channel is coming up. I loose SSH communication and even directly in console, the switch stop responding. This happens only with one the IE-2000. All five other work correctly.
Only po11 is causing the issue. I have no errors on the port and nothing in the logs.
Here are a few info for my setup:
2 x WS-C3850-48U, stacked running 16.3.7
6 x IE-2000-4TS-G-B running 15.2(4)EA5
Config on 3850
interface Port-channel11
description swSH6603
switchport mode trunk
switchport nonegotiate
ip arp inspection trust
spanning-tree portfast trunk
spanning-tree bpduguard disable
ip dhcp snooping trust
!
interface GigabitEthernet1/0/11
description swSH6603
switchport mode trunk
switchport nonegotiate
ip arp inspection trust
channel-protocol lacp
channel-group 11 mode active
spanning-tree portfast trunk
spanning-tree bpduguard disable
ip dhcp snooping trust
!
interface GigabitEthernet2/0/11
description swSH6603
switchport mode trunk
switchport nonegotiate
ip arp inspection trust
channel-protocol lacp
channel-group 11 mode active
spanning-tree portfast trunk
spanning-tree bpduguard disable
ip dhcp snooping trust
!
Group Port-channel Protocol Ports
------+-------------+-----------+-----------------------------------------------
11 Po11(SU) LACP Gi1/0/11(P) Gi2/0/11(P)
12 Po12(SU) LACP Gi1/0/12(P) Gi2/0/12(P)
13 Po13(SU) LACP Gi1/0/13(P) Gi2/0/13(P)
14 Po14(SD) LACP Gi1/0/14(P) Gi2/0/14(P)
15 Po15(SU) LACP Gi1/0/15(P) Gi2/0/15(P)
16 Po16(SU) LACP Gi1/0/16(P) Gi2/0/16(P)
Config on the IE-2000
interface Port-channel1
description vers swSH6601
switchport mode trunk
switchport nonegotiate
ip arp inspection trust
ip dhcp snooping trust
!
interface GigabitEthernet1/1
description vers swSH6601
switchport mode trunk
switchport nonegotiate
ip arp inspection trust
channel-protocol lacp
channel-group 1 mode active
ip dhcp snooping trust
!
interface GigabitEthernet1/2
description vers swSH6601
switchport mode trunk
switchport nonegotiate
ip arp inspection trust
channel-protocol lacp
channel-group 1 mode active
ip dhcp snooping trust
!
Group Port-channel Protocol Ports
------+-------------+-----------+-----------------------------------------------
1 Po1(SU) LACP Gi1/1(P) Gi1/2(P)
I already did several test. If I only bring one interface down, meaning that the port channel does not come down, I don't get the issue. Only when the entire port channel comes down and return to the UP state that I get the issue. I tried changing the physical ports on the switch, no luck. I tired changing the ports with one of the working IE-2000 and the issue followes the switch. I have triple checked the configuration on the problem switch with a working switch and, aside from their IP adresses, the configs are identical.
Any help would be appreciated.
Solved! Go to Solution.
12-04-2018 07:25 AM
Replaced the switch with a new one and the problem is gone. I will send the old switch back for a RMA.
11-27-2018 01:33 PM
11-28-2018 04:30 AM
Hi
Did you try to start up with a basic config, something like this? Then on if this works you can continue with the rest of the config.
interface GigabitEthernet X/Y/Z
switchport mode trunk
switchport nonegotiate
channel-protocol lacp
channel-group 11 mode active
BR
Gaston
11-29-2018 11:17 AM
Yes, I tried with no luck.
I have a new IE-2000 in order. I should receive it tomorrow. I will try and swap it. I will also try some new SFP and try with some other ethernet cables.
Hopefully. one of these is the issue.
11-29-2018 02:17 PM
Hello,
which SFPs are you using ? The below bugs applies to 3rd party SFP ACCELINK GLC-GE-100FX:
C3850-24S EtherChannel member keep down after opposite side(IE-2000) reboot
CSCvc61354
Description
Symptom:
There is a known quality issue with 3rd party SFP ACCELINK GLC-GE-100FX.
Conditions:
Issue was observed with the following topology:
IE2000(GLC-FE-100FX)=====C3850(GLC-GE-100FX)
IE2000(GLC-FE-100FX) connect with C3850(GLC-GE-100FX) with port channel(pagp).
When reboot IE2000 device, after IE2000 boot up, interface on C3850 may change portchannel status to I(Stand-along).
C3850 running 3.6.5E IOS.
IE2000 running 15.0(2)EA1 IOS.
Workaround:
Shut/no shut of the interface solves this issue.
11-30-2018 04:04 AM
I am using Cisco brand SFP.
I have a new IE-2000 at my desk. I will try and replace the one that is causing me issues on Tuesday next week. Hopefully, it will solve the issue.
Thanks.
12-04-2018 07:25 AM
Replaced the switch with a new one and the problem is gone. I will send the old switch back for a RMA.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide