LACP up with receive lacp-bpdu


Userlevel 6
Hi, all!

I heared that Cisco and Juniper switches have behavior (configure lacp) that ports working in stand-alone mode, but when receive lacp-bpdu then these ports became a LAG LACP.

Can we do this in Summit X770?

As I know in Summit LAG-LACP links is up only when receive lacp-bpdu.
Need to connect with servers.

Thank you!

5 replies

Userlevel 7
Hi Alexandr,
This feature, as I understand it to be implemented in a future EXOS release, is called LACP Fallback.
The LACP Fallback feature lets the user select a single port which will be automatically added to the aggregator if LACPDUs are not seen on any of the member ports within specified period of time.

At present, I can't provide a specific release version or date (sorry!), but I see that it's in the works. If there's something similar to this that already exists, I'm hopeful that someone will call it out here 🙂

-Drew
Userlevel 6
Drew C. wrote:

Hi Alexandr,
This feature, as I understand it to be implemented in a future EXOS release, is called LACP Fallback.
The LACP Fallback feature lets the user select a single port which will be automatically added to the aggregator if LACPDUs are not seen on any of the member ports within specified period of time.

At present, I can't provide a specific release version or date (sorry!), but I see that it's in the works. If there's something similar to this that already exists, I'm hopeful that someone will call it out here 🙂

-Drew

Can we try to use ACL with?
if - lacp-bpdu
then - run in CLI "ena sh 1 gr 1-2 "

Thank you!
Userlevel 6
HI Alexandr,

As Drew pointed out, this is in the roadmap for the future EXOS releases. Work-around can be implemented using ACL and UPM or clear-flow and ACL.

For the clear-flow solution, refer the following article:
https://gtacknowledge.extremenetworks.com/articles/How_To/LACP-configured-ports-interfere-with-PXE-DHCP-on-servers

For the UPM solution, refer the following article:

https://gtacknowledge.extremenetworks.com/articles/Solution/Work-around-for-LACP-fallback-backup-feature-support-in-EXOS

I have tested the UPM solution in my lab, it seems to work fine:

Sharing enabled on the connected device:
Load Sharing MonitorConfig Current Agg Ld Share Ld Share Agg Link Link Up
Master Master Control Algorithm Group Mbr State Transitions
==============================================================================
1:1 1:1 LACP L2 1:1 Y A 1
L2 1:2 Y A 1
==============================================================================
Link State: A-Active, D-Disabled, R-Ready, NP-Port not present, L-Loopback
Load Sharing Algorithm: (L2) Layer 2 address based, (L3) Layer 3 address based
(L3_L4) Layer 3 address and Layer 4 port based
(custom) User-selected address-based configuration
Custom Algorithm Configuration: ipv4 L3-and-L4, xor
Number of load sharing trunks: 1[/code]* Slot-1 Stack.9 # sh log10/07/2015 03:20:21.30 [i] Slot-1: Add port 1:2 to aggregator
10/07/2015 03:20:21.30 [i] Slot-1: Add port 1:1 to aggregator
10/07/2015 03:20:19.08 [i] Slot-1: Enabled_share_1:1
10/07/2015 03:20:18.94 Slot-1: Launched profile dynamiclacp for the event UPM Timer check_lacp_timer
Sharing disabled on the connected device:
* Slot-1 Stack.10 # sh log10/07/2015 03:22:19.16 [i] Slot-1: Disabled_share_1:1
10/07/2015 03:22:18.97 Slot-1: Launched profile dynamiclacp for the event UPM Timer check_lacp_timer
10/07/2015 03:21:38.93 Slot-1: Launched profile dynamiclacp for the event UPM Timer check_lacp_timer
10/07/2015 03:21:14.86 [i] Slot-1: Remove port 1:2 from aggregator
10/07/2015 03:21:14.86 [i] Slot-1: Remove port 1:1 from aggregator
* Slot-1 Stack.10 # sh sharing
Load Sharing Monitor
Config Current Agg Ld Share Ld Share Agg Link Link Up
Master Master Control Algorithm Group Mbr State Transitions
==============================================================================
==============================================================================
Link State: A-Active, D-Disabled, R-Ready, NP-Port not present, L-Loopback
Load Sharing Algorithm: (L2) Layer 2 address based, (L3) Layer 3 address based
(L3_L4) Layer 3 address and Layer 4 port based
(custom) User-selected address-based configuration
Custom Algorithm Configuration: ipv4 L3-and-L4, xor
Number of load sharing trunks: 0[/code][/code]I would suggest doing more tests before implementing in production.

Hope this helps!

P.S.
Thanks to Oscar Koot for both the solutions.
Userlevel 6
Prashanth KG wrote:

HI Alexandr,

As Drew pointed out, this is in the roadmap for the future EXOS releases. Work-around can be implemented using ACL and UPM or clear-flow and ACL.

For the clear-flow solution, refer the following article:
https://gtacknowledge.extremenetworks.com/articles/How_To/LACP-configured-ports-interfere-with-PXE-DHCP-on-servers

For the UPM solution, refer the following article:

https://gtacknowledge.extremenetworks.com/articles/Solution/Work-around-for-LACP-fallback-backup-feature-support-in-EXOS

I have tested the UPM solution in my lab, it seems to work fine:

Sharing enabled on the connected device:
Load Sharing MonitorConfig Current Agg Ld Share Ld Share Agg Link Link Up
Master Master Control Algorithm Group Mbr State Transitions
==============================================================================
1:1 1:1 LACP L2 1:1 Y A 1
L2 1:2 Y A 1
==============================================================================
Link State: A-Active, D-Disabled, R-Ready, NP-Port not present, L-Loopback
Load Sharing Algorithm: (L2) Layer 2 address based, (L3) Layer 3 address based
(L3_L4) Layer 3 address and Layer 4 port based
(custom) User-selected address-based configuration
Custom Algorithm Configuration: ipv4 L3-and-L4, xor
Number of load sharing trunks: 1[/code]* Slot-1 Stack.9 # sh log10/07/2015 03:20:21.30 [i] Slot-1: Add port 1:2 to aggregator
10/07/2015 03:20:21.30 [i] Slot-1: Add port 1:1 to aggregator
10/07/2015 03:20:19.08 [i] Slot-1: Enabled_share_1:1
10/07/2015 03:20:18.94 Slot-1: Launched profile dynamiclacp for the event UPM Timer check_lacp_timer
Sharing disabled on the connected device:
* Slot-1 Stack.10 # sh log10/07/2015 03:22:19.16 [i] Slot-1: Disabled_share_1:1
10/07/2015 03:22:18.97 Slot-1: Launched profile dynamiclacp for the event UPM Timer check_lacp_timer
10/07/2015 03:21:38.93 Slot-1: Launched profile dynamiclacp for the event UPM Timer check_lacp_timer
10/07/2015 03:21:14.86 [i] Slot-1: Remove port 1:2 from aggregator
10/07/2015 03:21:14.86 [i] Slot-1: Remove port 1:1 from aggregator
* Slot-1 Stack.10 # sh sharing
Load Sharing Monitor
Config Current Agg Ld Share Ld Share Agg Link Link Up
Master Master Control Algorithm Group Mbr State Transitions
==============================================================================
==============================================================================
Link State: A-Active, D-Disabled, R-Ready, NP-Port not present, L-Loopback
Load Sharing Algorithm: (L2) Layer 2 address based, (L3) Layer 3 address based
(L3_L4) Layer 3 address and Layer 4 port based
(custom) User-selected address-based configuration
Custom Algorithm Configuration: ipv4 L3-and-L4, xor
Number of load sharing trunks: 0[/code][/code]I would suggest doing more tests before implementing in production.

Hope this helps!

P.S.
Thanks to Oscar Koot for both the solutions.

Thank you a lot!!!
Userlevel 6
Hello, all!

FYI
LACP Fallback is appear in EXOS in 21.1 version.

Thank you!

Reply