problems with IRF: "heartbeat timed out"

[MichaelM55]

Hi,

I´d like to add a third A5500-HI switch to an existing IRF stack. (All with firmware release R5105)

So, let´s look at the setup from the HP A5500-HI installation manual:

 

Let´s concentrate on the connection netween "2" and "3"

1. renumbering the third switch to "3"

2. reboot

3. Plugin the DAC-Cable between 2 and 3

4. On switch "2"

interface Ten-GigabitEthernet 2/1/2
shutdown
irf-port 2/1
port group interface Ten-GigabitEthernet 2/1/2 
interface Ten-GigabitEthernet 2/1/2
undo shutdown
save

 5. On switch "3":

interface Ten-GigabitEthernet 3/2/1
shutdown
irf-port 3/2
port group interface Ten-GigabitEthernet 3/2/1 
interface Ten-GigabitEthernet 3/2/1
undo shutdown
save
y

 6. After that on switch "3" let´s do a

irf-port-configuration active

 6.1 "3" doesn´t reboot at all. but output on "3" says:

#May 10 08:45:27:326 2000 HP STM/4/LINK STATUS CHANGE:
 Trap 1.3.6.1.4.1.25506.2.91.6.0.1: Physical index of the member is 65536, member ID is 3. Link status of the IRF port with port index 2 and member ID 3 turned to 1.
%May 10 08:45:27:607 2000 HP STM/6/STM_LINK_STATUS_UP:
 IRF port 2 is up.
#May 10 08:45:47:503 2000 HP STM/4/LINK STATUS CHANGE:
 Trap 1.3.6.1.4.1.25506.2.91.6.0.1: Physical index of the member is 65536, member ID is 3. Link status of the IRF port with port index 2 and member ID 3 turned to 3.
#May 10 08:45:47:784 2000 HP STM/4/LINK STATUS CHANGE:
 Trap 1.3.6.1.4.1.25506.2.91.6.0.1: Physical index of the member is 65536, member ID is 3. Link status of the IRF port with port index 2 and member ID 3 turned to 2.
%May 10 08:45:48:070 2000 HP STM/2/STM_LINK_STATUS_TIMEOUT:
 IRF port 2 is down because heartbeat timed out.
%May 10 08:45:48:211 2000 HP STM/3/STM_LINK_STATUS_DOWN:
 IRF port 2 is down.

 7. Let´s look on "2" with

dis log rev

 7.1 output of "2":

%Sep 10 07:43:09:685 2000 HP DRVDBG/2/LOG Info: -Slot=2;
 The port Ten2/1/2 can't receive irf pkt, please check.
%Sep 10 07:43:00:657 2000 HP DRVDBG/2/LOG Info: -Slot=2;
 The port Ten2/1/2 can't receive irf pkt, please check.
%Sep 10 07:42:51:627 2000 HP DRVDBG/2/LOG Info: -Slot=2;
 The port Ten2/1/2 can't receive irf pkt, please check.
%Sep 10 07:42:42:622 2000 HP DRVDBG/2/LOG Info: -Slot=2;
 The port Ten2/1/2 can't receive irf pkt, please check.
%Sep 10 07:42:33:617 2000 HP DRVDBG/2/LOG Info: -Slot=2;
 The port Ten2/1/2 can't receive irf pkt, please check.
%Sep 10 07:42:25:222 2000 HP IFNET/3/LINK_UPDOWN: Ten-GigabitEthernet2/1/2 link status is UP.
%Sep 10 07:39:56:252 2000 HP IFNET/3/LINK_UPDOWN: Ten-GigabitEthernet2/1/2 link status is DOWN.

I did this several times, but nothing changed, delete config on "3", reboot, do it again. Perhaps some kind of fallacy?

[Peter_Debruyne · Accepted solution]

Most features on the comware switches are active when you execute the cli command.

There are (AFAIK) 2 exceptions:

* MSTP region configuration (both initial config and later updates)

* IRF Port configuration (both initial config and later updates)

In this case, the new switch IRF config was done and activated with the "irf-port-configuration active" (OK), the existing switch config was updated, but the new config was not activated.

This means that the exising IRF system was not running the IRF protocol on the new ports, resulting in the "heartbeat time out" message on the switch which is trying to join.

Simply executing the "irf-port-configuration active" on the existing IRF system should resolve the issue,

Hope this helps,Peter

[BoT]

Thank you for posting this problem fix.
I just had exactly the above descripted problem and got it fix just after reading this :-)