Description

This article explains "ORE not alive" meaning and fix.

Symptoms

On Dual RE platform, following outputs might be seen in /var/log/mastership periodically.

 

Oct 24 01:14:45 mcontrol_ore_alive_set: other RE is not alive

Oct 24 01:14:45 ORE not alive, keepalive loss 1 above threshold of 0

Oct 24 01:14:45 mcontrol_ore_alive_set: other RE is alive

Oct 24 04:03:58 mcontrol_ore_alive_set: other RE is not alive

Oct 24 04:03:58 ORE not alive, keepalive loss 1 above threshold of 0

Oct 24 04:04:00 event = E_NO_IPC, state = master, param = 0x0x0

Oct 24 04:04:00 currentAction = A_WARN1

 

Oct 24 04:04:00 No response from the other routing engine for the last 2 seconds.

 

Oct 24 04:04:00 Currentstate master NextState master reason_code 1

Oct 24 04:04:00 new state = master

Oct 24 04:04:00 vc master RE state change: in-synch -> initializing

Oct 24 04:04:00 vc master RE append vc ext disabled

Oct 24 04:04:00 mcontrol_ore_alive_set: other RE is alive

Oct 24 04:04:00 event = E_ORE_B, state = master, param = 0x0xc17e008

Oct 24 04:04:00 currentAction = A_NOOP

 

Oct 24 04:04:00 Currentstate master NextState master reason_code 1

Oct 24 04:04:00 new state = master

Oct 24 04:04:00 vc master RE state change: initializing -> synching

Oct 24 04:04:00 vc master RE append vc ext enabled

Oct 24 04:04:00 vc master RE recv backup RE vc data, mid=255 sn=???????????? slots=8

Oct 24 04:04:00 vc master RE state change: synching -> in-synch

Solution

On Dual RE platform, keepalive is sent every second between master RE and backup RE to confirm master aliveness.

The syslog is caused by missing keepalive from backup RE in a short period.


There should be no harm caused by the missing keepalive.

Normally, the keepalive will be resumed soon.


PR1689702 fixed the issue by increasing detection period to avoid false alarm.

Modification History

2024-11-07 : Article Created