LACP PROBLEM

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Announcements
Please sign in to see details of an important advisory in our Customer Advisories area.

LACP PROBLEM

L1 Bithead

Hi all,

 

I have some problems with LACP. Sometimes, randomly, the interfaces move out of AE-group.

I can see from log this error message: "receive PDU partner does not match local actor ".

Below the file l2ctrld.log .

 

 

 

 

2019-09-17 23:19:54.588 +0200 ethernet1/10 idx 73 received pdu partner does not match local actor
2019-09-17 23:19:54.588 +0200 Recved LACPDU actor:
  sys_pri 32768, system_mac X, key 22, port_pri 32768, port_num 10245, state 0x47
partner
  sys_pri 0, system_mac X idx 73 received pdu partner does not match local actor
2019-09-17 23:19:54.588 +0200 Recved LACPDU actor:
  sys_pri 32768, system_mac X, key 22, port_pri 32768, port_num 10245, state 0x47
partner
  sys_pri 32768, system_mac X, key 16, port_pri 32768, port_num 73, state 0x2
2019-09-17 23:20:00.338 +0200 ethernet1/10 idx 73, mux state change ATTACHED=>RX_TX
2019-09-17 23:20:00.338 +0200 ethernet1/10 idx 73, mux state in RX_TX
2019-09-17 23:20:00.338 +0200 post LACP event to DP: if_idx 73, up 1
2019-09-17 23:20:00.338 +0200 log ethernet1/10 idx 73 join lag
2019-09-17 23:20:09.838 +0200 Got port 73 event, link 0, speed 4, duplex 2
2019-09-17 23:20:09.838 +0200 port 73 is full duplex. Enable lacp
2019-09-17 23:20:09.838 +0200 ethernet1/10 idx 73 mux state change RX_TX=>ATTACHED, select_state Unselected(Link down), partner state 0x3f
2019-09-17 23:20:09.838 +0200 post LACP event to DP: if_idx 73, up 0
2019-09-17 23:20:09.838 +0200 log ethernet1/10 idx 73 leaves lag. sel state Unselected(Link down)
2019-09-17 23:20:10.088 +0200 ethernet1/10 idx 73, mux state change ATTACHED=>DETACHED
2019-09-17 23:20:13.087 +0200 ethernet1/10 idx 73, current_while expired.
2019-09-17 23:20:13.087 +0200 ethernet1/10 idx 73, rx state change CURRENT=>EXPIRED
2019-09-17 23:20:16.087 +0200 ethernet1/10 idx 73, current_while expired.
2019-09-17 23:20:16.087 +0200 ethernet1/10 idx 73, rx state change EXPIRED=>DEFAULTED
2019-09-17 23:20:42.119 +0200 ethernet1/8 idx 71 set to unselected, clear actor sync
2019-09-17 23:20:42.338 +0200 ethernet1/8 idx 71 mux state change RX_TX=>ATTACHED, select_state Unselected(Negotiation failed), partner state 0x3
2019-09-17 23:20:42.338 +0200 post LACP event to DP: if_idx 71, up 0
2019-09-17 23:20:42.338 +0200 log ethernet1/8 idx 71 leaves lag. sel state Unselected(Negotiation failed)
2019-09-17 23:20:42.588 +0200 Got port 71 event, link 0, speed 4, duplex 2
2019-09-17 23:20:42.588 +0200 port 71 is full duplex. Enable lacp
2019-09-17 23:20:42.588 +0200 ethernet1/8 idx 71, mux state change ATTACHED=>DETACHED
2019-09-17 23:20:46.087 +0200 ethernet1/8 idx 71, current_while expired.
2019-09-17 23:20:46.087 +0200 ethernet1/8 idx 71, rx state change CURRENT=>EXPIRED
2019-09-17 23:20:49.087 +0200 ethernet1/8 idx 71, current_while expired.
2019-09-17 23:20:49.087 +0200 ethernet1/8 idx 71, rx state change EXPIRED=>DEFAULTED
2019-09-17 23:21:34.142 +0200 ethernet1/20 idx 83 received pdu partner does not match local actor
2019-09-17 23:21:34.143 +0200 Recved LACPDU actor:
  sys_pri 127, system_mac X, key 2, port_pri 127, port_num 2, state 0x47
partner
  sys_pri 1, system_mac 00:00:00:00:00:00, key 2, port_pri 1, port_num 2, state 0x46
2019-09-17 23:21:34.338 +0200 ethernet1/20 idx 83 mux state change RX_TX=>ATTACHED, select_state Selected, partner state 0x47
2019-09-17 23:21:34.338 +0200 post LACP event to DP: if_idx 83, up 0
2019-09-17 23:21:34.338 +0200 log ethernet1/20 idx 83 leaves lag. sel state Selected
2019-09-17 23:21:36.587 +0200 ethernet1/20 idx 83, mux state change ATTACHED=>RX_TX
2019-09-17 23:21:36.588 +0200 ethernet1/20 idx 83, mux state in RX_TX
2019-09-17 23:21:36.588 +0200 post LACP event to DP: if_idx 83, up 1
2019-09-17 23:21:36.588 +0200 log ethernet1/20 idx 83 join lag
2019-09-17 23:24:35.087 +0200 log ethernet1/7 idx 70 leaves lag. sel state Unselected(Link down)
2019-09-17 23:24:42.087 +0200 log ethernet1/9 idx 72 leaves lag. sel state Unselected(Link down)
2019-09-17 23:25:09.087 +0200 log ethernet1/10 idx 73 leaves lag. sel state Unselected(Link down)
2019-09-17 23:25:42.087 +0200 log ethernet1/8 idx 71 leaves lag. sel state Unselected(Link down)
2019-09-17 23:29:35.087 +0200 log ethernet1/7 idx 70 leaves lag. sel state Unselected(Link down)
2019-09-17 23:29:42.087 +0200 log ethernet1/9 idx 72 leaves lag. sel state Unselected(Link down)
2019-09-17 23:30:09.087 +0200 log ethernet1/10 idx 73 leaves lag. sel state Unselected(Link down)
2019-09-17 23:30:42.087 +0200 log ethernet1/8 idx 71 leaves lag. sel state Unselected(Link down)
2019-09-17 23:34:35.087 +0200 log ethernet1/7 idx 70 leaves lag. sel state Unselected(Link down)

 

 

 

 

 

 

 

Could you help me?

THX

 

CHANGÒ
2 accepted solutions

Accepted Solutions

L1 Bithead

Hi @Chango ,

 

Most probably one interface from aggregate group is connected to one switch and other to 2nd switch and both the physical switches are virtually clustered into one.

When we do this on switch it will generate one system ID which would be virtual and will use it for lacp negotiation ( it will not use physical system ID since it will be two in numbers and each would be specific to each switch.

2019/09/27 16:10:17 critical lacp    ethern lacp-up 0  LACP interface ethernet1/7 moved into AE-group ae1.

2019/09/27 16:10:16 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/9 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:13 critical lacp    ethern lacp-up 0  LACP interface ethernet1/9 moved into AE-group ae1.

2019/09/27 16:10:11 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/8 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:08 critical lacp    ethern lacp-up 0  LACP interface ethernet1/8 moved into AE-group ae1.

2019/09/27 16:10:06 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/22 chassis id: 44:ec:ce:64:c9:12. New neighbor added to mib.

2019/09/27 16:10:01 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/10 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:01 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/21 chassis id: 44:ec:ce:6f:81:28. New neighbor added to mib.

 

 am seeing that the aggregate group (ae1) got the actor's virtual mac but it is flapping because peer is configured on fast rate and firewall is requesting for the next packet again in few seconds.

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 32768, system_mac 02:00:00:00:00:64, key 22, port_pri 32768, port_num 6149, state 0x7f

mp l2ctrld.log 2019-09-27 16:10:06 partner

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 0, system_mac 00:00:00:00:00:00, key 0, port_pri 0, port_num 0, state 0x0

 

 

****In fast mode if we don't receive lacpdu for 3 seconds lacp will go down. So it's very sensitive to network churn**

 

Please configure both the devices on slow transmission rate .

 

Regards.

EC

View solution in original post

L1 Bithead

Hi @Chango ,

 

Most probably one interface from aggregate group is connected to one switch and other to 2nd switch and both the physical switches are virtually clustered into one.

When we do this on switch it will generate one system ID which would be virtual and will use it for lacp negotiation ( it will not use physical system ID since it will be two in numbers and each would be specific to each switch.

2019/09/27 16:10:17 critical lacp ethern lacp-up 0 LACP interface ethernet1/7 moved into AE-group ae1.

2019/09/27 16:10:16 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/9 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:13 critical lacp ethern lacp-up 0 LACP interface ethernet1/9 moved into AE-group ae1.

2019/09/27 16:10:11 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/8 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:08 critical lacp ethern lacp-up 0 LACP interface ethernet1/8 moved into AE-group ae1.

2019/09/27 16:10:06 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/22 chassis id: 44:ec:ce:64:c9:12. New neighbor added to mib.

2019/09/27 16:10:01 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/10 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:01 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/21 chassis id: 44:ec:ce:6f:81:28. New neighbor added to mib.

 

am seeing that the aggregate group (ae1) got the actor's virtual mac but it is flapping because peer is configured on fast rate and firewall is requesting for the next packet again in few seconds.

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 32768, system_mac 02:00:00:00:00:64, key 22, port_pri 32768, port_num 6149, state 0x7f

mp l2ctrld.log 2019-09-27 16:10:06 partner

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 0, system_mac 00:00:00:00:00:00, key 0, port_pri 0, port_num 0, state 0x0

 

 

****In fast mode if we don't receive lacpdu for 3 seconds lacp will go down. So it's very sensitive to network churn**

 

Please configure both the devices on slow transmission rate .

 

Regards.
EC

CE

View solution in original post

10 REPLIES 10

L3 Networker

Was the physical layer issue ruled out (cables, ports, SFPs)?

Yes.

I have 2 aggregate:

1 with Juniper and 1 with Cisco.

The aggregates are configured in active(Juniper or Cisco) passive(PA) and trasmission rate setted to fast.

I have problems on both aggregates.

 

CHANGÒ

Have you tried changing one AE group to "slow" mode to see in that stabilise the situation? Maybe fast is too aggressive, and occasionally you have some packet dops etc. 

@myky 

I can try but i have PA-5260 end i want the fast mode.

However on Cisco peer i have different partner's number (6148,6149,10244 and 10245), on Juniper i have the same partner's number(4), is it correct?

Another question, the port priority must be the same ?

immagine-2.png

CHANGÒ

i have 5220 connected to switch

one end is fast and other is slow

 

 

MP

Help the community: Like helpful comments and mark solutions.

@MP18 

Which is setted to slow? Firewall?

What is the switch model?

CHANGÒ

L1 Bithead

Hi @Chango ,

 

Most probably one interface from aggregate group is connected to one switch and other to 2nd switch and both the physical switches are virtually clustered into one.

When we do this on switch it will generate one system ID which would be virtual and will use it for lacp negotiation ( it will not use physical system ID since it will be two in numbers and each would be specific to each switch.

2019/09/27 16:10:17 critical lacp    ethern lacp-up 0  LACP interface ethernet1/7 moved into AE-group ae1.

2019/09/27 16:10:16 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/9 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:13 critical lacp    ethern lacp-up 0  LACP interface ethernet1/9 moved into AE-group ae1.

2019/09/27 16:10:11 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/8 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:08 critical lacp    ethern lacp-up 0  LACP interface ethernet1/8 moved into AE-group ae1.

2019/09/27 16:10:06 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/22 chassis id: 44:ec:ce:64:c9:12. New neighbor added to mib.

2019/09/27 16:10:01 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/10 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:01 info     lldp    ethern mib cha 0  Update: Received change on local interface ethernet1/21 chassis id: 44:ec:ce:6f:81:28. New neighbor added to mib.

 

 am seeing that the aggregate group (ae1) got the actor's virtual mac but it is flapping because peer is configured on fast rate and firewall is requesting for the next packet again in few seconds.

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 32768, system_mac 02:00:00:00:00:64, key 22, port_pri 32768, port_num 6149, state 0x7f

mp l2ctrld.log 2019-09-27 16:10:06 partner

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 0, system_mac 00:00:00:00:00:00, key 0, port_pri 0, port_num 0, state 0x0

 

 

****In fast mode if we don't receive lacpdu for 3 seconds lacp will go down. So it's very sensitive to network churn**

 

Please configure both the devices on slow transmission rate .

 

Regards.

EC

L1 Bithead

Hi @Chango ,

 

Most probably one interface from aggregate group is connected to one switch and other to 2nd switch and both the physical switches are virtually clustered into one.

When we do this on switch it will generate one system ID which would be virtual and will use it for lacp negotiation ( it will not use physical system ID since it will be two in numbers and each would be specific to each switch.

2019/09/27 16:10:17 critical lacp ethern lacp-up 0 LACP interface ethernet1/7 moved into AE-group ae1.

2019/09/27 16:10:16 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/9 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:13 critical lacp ethern lacp-up 0 LACP interface ethernet1/9 moved into AE-group ae1.

2019/09/27 16:10:11 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/8 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:08 critical lacp ethern lacp-up 0 LACP interface ethernet1/8 moved into AE-group ae1.

2019/09/27 16:10:06 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/22 chassis id: 44:ec:ce:64:c9:12. New neighbor added to mib.

2019/09/27 16:10:01 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/10 chassis id: d8:24:bd:91:fd:c0. New neighbor added to mib.

2019/09/27 16:10:01 info lldp ethern mib cha 0 Update: Received change on local interface ethernet1/21 chassis id: 44:ec:ce:6f:81:28. New neighbor added to mib.

 

am seeing that the aggregate group (ae1) got the actor's virtual mac but it is flapping because peer is configured on fast rate and firewall is requesting for the next packet again in few seconds.

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 32768, system_mac 02:00:00:00:00:64, key 22, port_pri 32768, port_num 6149, state 0x7f

mp l2ctrld.log 2019-09-27 16:10:06 partner

 

mp l2ctrld.log 2019-09-27 16:10:06 sys_pri 0, system_mac 00:00:00:00:00:00, key 0, port_pri 0, port_num 0, state 0x0

 

 

****In fast mode if we don't receive lacpdu for 3 seconds lacp will go down. So it's very sensitive to network churn**

 

Please configure both the devices on slow transmission rate .

 

Regards.
EC

CE

I have the same issue with PA-3220 the LACP transmission rate is set to slow on PA-Firewalls. I face this issue during migration from Old PA3020 to PA3220. This is the error message

 

SYSTEM ALERT : critical : LACP interface ethernet1/1 moved out of AE-group ae1. Selection state Unselected(Link down)

Any update on resolution?

  • 2 accepted solutions
  • 34050 Views
  • 10 replies
  • 1 Likes
Like what you see?

Show your appreciation!

Click Like if a post is helpful to you or if you just want to show your support.

Click Accept as Solution to acknowledge that the answer to your question has been provided.

The button appears next to the replies on topics you’ve started. The member who gave the solution and all future visitors to this topic will appreciate it!

These simple actions take just seconds of your time, but go a long way in showing appreciation for community members and the LIVEcommunity as a whole!

The LIVEcommunity thanks you for your participation!