5400 Switches with ETAP architecture

[zzl630]

update:
We decided for our architecture REP is not required. We made a test setup with redundant controller racks, switchs, ETAPS, AENTRs. We set up the redundant switches as supervisors and EN2TRs as backup supervisors. Having trouble as supervisors keep on switching between each other(see below logs). The original plan was to set the EN2TR as supervisor, thinking may need to go back.

Feb 23 11:51:41.152 UTC: %UDLD-4-UDLD_PORT_DISABLED: UDLD disabled interface Gi1/1, aggressive mode failure detected
Feb 23 11:51:41.152 UTC: %PM-4-ERR_DISABLE: udld error detected on Gi1/1, putting Gi1/1 in err-disable state
Feb 23 11:51:42.155 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet1/1, changed state to down
Feb 23 11:51:43.155 UTC: %LINK-3-UPDOWN: Interface GigabitEthernet1/1, changed state to down
Feb 23 11:52:11.149 UTC: %PM-4-ERR_RECOVER: Attempting to recover from udld err-disable state on Gi1/1
Feb 23 11:52:13.187 UTC: %LINK-3-UPDOWN: Interface GigabitEthernet1/1, changed state to up
Feb 23 11:52:13.187 UTC: %DLR-6-SUPSTATECHG: Ring 1: The device is now the ACTIVE supervisor
Feb 23 11:52:13.191 UTC: %DLR-6-RINGNORMAL: Ring 1: The DLR ring is now in NORMAL state
Feb 23 11:52:13.191 UTC: %DLR-2-RINGFAULT: Ring 1: The DLR ring is now in FAULT state
Feb 23 11:52:13.194 UTC: %DLR-2-SUPPARTIALFAULT: Ring 1: The supervisor device is now in PARTIAL FAULT state
Feb 23 11:52:14.190 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet1/1, changed state to up
Feb 23 11:52:30.647 UTC: %DLR-6-SUPSTATECHG: Ring 1: The device is now the ACTIVE supervisor
Feb 23 11:52:30.647 UTC: %DLR-6-RINGNORMAL: Ring 1: The DLR ring is now in NORMAL state
Feb 23 11:52:30.650 UTC: %DLR-2-RINGFAULT: Ring 1: The DLR ring is now in FAULT state
Feb 23 11:52:30.650 UTC: %DLR-6-SUPSTATECHG: Ring 1: The device is now the ACTIVE supervisor
Feb 23 11:52:30.650 UTC: %DLR-6-RINGNORMAL: Ring 1: The DLR ring is now in NORMAL state
Feb 23 11:52:30.654 UTC: %DLR-2-RINGFAULT: Ring 1: The DLR ring is now in FAULT state
Feb 23 11:52:30.657 UTC: %DLR-2-SUPPARTIALFAULT: Ring 1: The supervisor device is now in PARTIAL FAULT state
Feb 23 11:52:40.679 UTC: %DLR-6-SUPSTATECHG: Ring 1: The device is now the ACTIVE supervisor
Feb 23 11:52:40.679 UTC: %DLR-6-RINGNORMAL: Ring 1: The DLR ring is now in NORMAL state
Feb 23 11:52:40.682 UTC: %DLR-2-RINGFAULT: Ring 1: The DLR ring is now in FAULT state
Feb 23 11:52:40.682 UTC: %DLR-2-SUPPARTIALFAULT: Ring 1: The supervisor device is now in PARTIAL FAULT state
Feb 23 11:53:17.510 UTC: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet1/6, changed state to down
Feb 23 11:53:18.513 UTC: %LINK-3-UPDOWN: Interface FastEthernet1/6, changed state to down

 

[Contr_Conn]

Hard to say what is happening without knowing exact configuration and Stratix settings. If no REP, then why you need Stratix?
All below just a guess based on my last 8 years living mostly in DLR/Redundancy world
It feels like you have a Stratix loop outside of the DLR, redundant Gateway may be needed?
Since you have many fiber nodes, you may have non-DLR fiber converters. Even one converter will cause similar instability.
DLR ports on Stratix switches must be configured to Multiport Automation Device.
Also, with multiple Fiber segments, you need to adjust Beacon interval and timeouts.
Start with small, no fiber DLR, get it stable, SYNC redundancy, then keep adding nodes and segments.

 

[zzl630]

Exactly. Previous project went over-engineering and unfortunately some lesson learns just passed to us.
We started with simple set up with just DLR including just two switches, then switches with redundant controllers, then switches with etaps and remote IO.I was hoping get the picture setup tested before going full setup in the field.

 

[Contr_Conn]

What else Stratix 5400 do besides to be an expensive media converter?
Are there any other connections to these?
Based on your log, I think it may be a connection from switch to switch, possibly via ann upstream switch, and this will create an unmanaged loop.

 

[zzl630]

It will be connected to an upper level switch then it is accessible for an EWS.

I don't see an unmanaged loop unless my laptop connected to one ETAP and 2 other computers are connected to a switch is causing any issue.

 

[Contr_Conn]

If switches will be connected to an upper level switch, then they must be configured as Redundant gateways. As a result no other devices should be connected to ports other than DLR and uplinks.
Once you have Redundant Gateways, then you can’t use ETAPs they are not compatible with Redundant Gateways.
Check publication ENET-AT007 for more details.