Problems with LACP trunking Core 8212zl - Access 2910al

[xconsole]

Two Coreswitches (8212zl) connected via distributed trunking are working together.

All Access-Switches (2910al) are connected in the same configuration.

! Access-Switch
; J9148A Configuration Editor; Created on release #W.15.08.0007
; Ver #02:11.05:16

trunk 47-48 trk1 lacp

! Core-Switces
; J9091A Configuration Editor; Created on release #K.15.07.0008
; Ver #02:1b.2f:36

switch-interconnect Trk1
trunk C1,I1 Trk1 Trunk
trunk A1 Trk10 LACP
trunk A3 Trk11 LACP
trunk A5 Trk12 LACP

Everything runs fine, but after an failure-test (we take out the active CPU - the secondary started) all LACP-Links are in "Failure" Status.

         LACP      Trunk     Port                LACP      Admin   Oper
   Port   Enabled   Group     Status    Partner   Status    Key     Key
   ----   -------   -------   -------   -------   -------   ------  ------
   A1     Active    Trk10     Blocked   Yes       Failure   0        299   
   A3     Active    Trk11     Blocked   Yes       Failure   0        300   
   A5     Active    Trk12     Blocked   Yes       Failure   0        301   

After disable and enable the links became activ.

        LACP      Trunk     Port                LACP      Admin   Oper
   Port   Enabled   Group     Status    Partner   Status    Key     Key
   ----   -------   -------   -------   -------   -------   ------  ------
   A1     Active    Trk10     Up        Yes       Success   0        299   
   A3     Active    Trk11     Up        Yes       Success   0        300   
   A5     Active    Trk12     Up        Yes       Success   0        301  

Why did this run automatic? What is happend? Did we anything forgotten?

Thank a lot!

Matthias

[Fredrik Lönnman · Accepted solution]

Shouldn't it be dt-lacp against the access switches?

trunk A1 trk10 dt-lacp
trunk A3 trk11 dt-lacp
trunk A5 trk12 dt-lacp