JUNOS High Availability

Final Burn-In Check

As a final test, if possible, you should power down and then power up the router again, and leave it running for at least 48 hours. After this period, check the system logs and boards status.

Power Down the Router

To power down the router, first stop the backup RE:

lab@r1-re0> request routing-engine login other-routing-engine
lab@r1-re1> request system halt

After a few seconds, the backup RE will stop, and your Telnet session will fall back to the master RE. After this, stop the master RE:

lab@r1-re0> request system halt

Both REs are now halted. Turn the power off.

Power On the Router/Burn-In Test

Connect a console cable to RE 0 (capturing any messages on file), and turn the power on. Check for any active alarms and the status of the router boards. Check the content of the messages log for any error with the show log messages command.

The router must be left powered on for one or two days.

Final Checks and Power Down

Be sure to check that the uptime for all the FPCs and SFMs is about the same. If it is not, this means one of the boards has restarted during the burn-in period—check in the logs for details. Here is sample output from an M10i router:

lab@r1-re0> show chassis fpc detail Slot 0 information: State Online Logical slot 0 Temperature 34 degrees C / 93 degrees F Total CPU DRAM 8 MB Total SRAM 1 MB Total SDRAM 128 MB Total notification SDRAM 24 MB I/O Manager ASIC information Version 2.0, Foundry IBM, Part number 0 Start time: 2009-03-19 10:15:24 CET Uptime: 2 days, ...