Installation Guide

Table Of Contents
To know the presence of MP by LP, accounting of IPC and keep-alive messages from the MP are performed in the LP.
The current state of MP presence is displayed when you enter the ipc show mp-presence state CLI command in LP. The command
output is as follows:
LP-2#ipc show mp-presence state
MP-PRESENCE is in MP_ALIVE
The following table lists all the MP-Presence states maintained by LP.
TABLE 40 MP-Presence States for Show MP-Presence State
MP Presence State
MP_ALIVE MP_WAIT
MP_SWITCH MP_ALIVE_DISABLED
By default, MP Presence is enabled. MP presence is toggled when you enter the ipc toggle mp-presence CLI command in LP. The
command output is as follows:
LP-2#ipc toggle mp-presence
IPC MP Presence disabled
LP-2#ipc toggle mp-presence
IPC MP Presence enabled
The threshold time is set in the chassis when you enter the command dm mp-presence-threshold-time . This command displays a
warning message in the LP console before LP is reset after nding the Active MP.
The allowed threshold time is 0 to 35 seconds only. The maximum limit of 35 seconds is maintained because the LP is reset on the MP
absence in 35 seconds. The input value 0 clears mp-presence-threshold-time .
Commands for setting mp-presence-threshold time are shown in the following table.
TABLE 41 Operations for Setting mp-presence-threshold Time
Operation Result
LP-1#dm mp-presence-threshold-time 10 MP Presence threshold time is set to 10 seconds.
LP-1#dm mp-presence-threshold-time 0 MP Presence threshold time is cleared.
When LP detects no MPs are present, LP will reset in another 35 seconds. The below warning message is displayed in LP console
before reset.
WARNING
mp_presence_threshold_log: LP will go for a reset in another 35 seconds.
Rolling Reboot
The Rolling Reboot feature provides a solution for continuously rebooting the line card when a system failure occurs from causes
including BIST failure, ECC errors, module init failure, and hardware failure during system initialization. When there is failure, the system
automatically reboots in an innite loop until the failure is resolved. Repetitious reboot cycles consume CPU resources, power, and result
in unwanted IPC trac to the MP.
When the system detects the continuous reboot of a line card, that line card is placed into interactive or down state mode after ten (10)
consecutive reboot cycles, and a failure message is sent to the MP.
Managing the device
Extreme Routing MLX Series Hardware Installation Guide
208 53-1004203-04