Communication across BAGG not working

[Syncorp12]

I am trying to connect 2 A5500-48G-POE+ switches via BAGG trunk ports and I am unable to communicate between the two.  I can see the LLDP neighbor information between the 2 but cannot ping between them using the vlan5 ip's.

RD-5500-IDF - Production switch uplinked to our core
RD-SPARESW-1 - Switch being configured as standby

RD-5500-IDF config:
vlan 5
description MGMT-10.1.5.0/24
!
interface Vlan-interface5
ip address 10.1.5.16 255.255.255.0
!
interface Bridge-Aggregation2
port link-type trunk
undo port trunk permit vlan 1
port trunk permit vlan 2 to 3 5 200 to 203 255
port trunk pvid vlan 5
link-aggregation mode dynamic
!
interface GigabitEthernet1/0/45
port link-mode bridge
description TEMP TRUNK TO SPARESW
port link-type trunk
undo port trunk permit vlan 1
port trunk permit vlan 2 to 3 5 200 to 203 241 255
port trunk pvid vlan 5
qos trust dscp
port link-aggregation group 2

RD-SPARESW-1 config:
vlan 5
description MGMT-10.1.5.0/24
!
interface Vlan-interface5
ip address 10.1.5.6 255.255.255.0
!
interface Bridge-Aggregation1
description Switch Interconnect
port link-type trunk
undo port trunk permit vlan 1
port trunk permit vlan 2 to 3 5 200 to 203 255
port trunk pvid vlan 5
link-aggregation mode dynamic
!
interface GigabitEthernet1/0/48
port link-mode bridge
description Switch Interconnect
port link-type trunk
undo port trunk permit vlan 1
port trunk permit vlan 2 to 3 5 200 to 203 255
port trunk pvid vlan 5
qos trust dscp
port link-aggregation group 1

[parnassus]

It's essential that BAGG member links are co-terminus. Are they?

[Syncorp12]

The links have been setup using 2 interfaces on each swith as follows:

RD-5500-IDF gi1/0/45 --- RD-SPARESW-1 gi1/0/48
RD-5500-IDF gi1/0/46 --- RD-SPARESW-1 gi2/0/48

RD-SPARESW-1 is a stack. I have tried shutting down the second switch in the stack and theres still no IP communication. All ports are configured the same.

[Syncorp12]

If I remove the BAGG config from the physical interface I am able to communiucate between the two switches.

[parnassus · Accepted solution]

Hi, I could have written a long post but I'm in a hurry now so I can't just now...but...please check this thread:

https://community.hpe.com/t5/Comware-Based/Link-Aggregation/m-p/6992974/highlight/true#M10257

As written above...provide the BAGG status on both ends with the command:

display link-aggregation verbose

that's to start understaning what went wrong (and why).

Edit: I suppose the stack you're referring on is an well formed IRF.

[Syncorp12]

Thank you! I deleted and rebuilt the BAGG as instructed and it is working fine now. I didn't realize how picky the configuration steps were in exicution.

Have a great day! 

[parnassus]

Great! Glad you fixed it.

[VoIP-Buddy]

It is very picky.  You do the absolute minimum on the member interfaces... connecting it to the BAGG.  Then all of the configuration work is done via the BAGG interface.  THe configuration is then cascaded down to the member ports. 

Interestingly I discovered last week that you cannot do a qos trust dscp at the BAGG level but you still can at the member interface level!  It is also allowed!  The member interfaces will not unselect when you do that.

Regards,

David