This article describes the issue of being unable to save the rescue configuration on a clustered pair of 5800s. Each node has two REs installed (to take advantage of a second control link).

 When a second RE is present and an attempt is made to save the rescue configuration, the following response is received from the CLI:

{{primary:node0}[edit]
root@ndc-sec0# run request system configuration rescue save
error: Could not connect to unknown : No route to host


This is merely a superficial issue within JUNOS and does not indicate a problem. The rescue configuration is being saved and there is a convenient way to validate this from the CLI. The rescue configuration gets saved to the /config directory. You can run a file list command with the detail tag to see when the rescue file was last modified.

Here, the rescue save command was executed, a system uptime was run to show the current time within Junos, and the file list detail command shows the file present with the current timestamp.

{primary:node0}[edit]
root@ndc-sec0# run request system configuration rescue save
error: Could not connect to unknown : No route to host

{primary:node0}[edit]
root@ndc-sec0# run show system uptime
node0:
--------------------------------------------------------------------------
Current time: 2012-03-24 16:12:39 UTC
System booted: 2012-03-22 16:59:54 UTC (1d 23:12 ago)
Protocols started: 2012-03-22 17:02:58 UTC (1d 23:09 ago)
Last configured: 2012-03-22 17:22:20 UTC (1d 22:50 ago) by root
4:12PM up 1 day, 23:13, 1 user, load averages: 0.05, 0.03, 0.00

{primary:node0}[edit]
root@ndc-sec0# run file list /config/rescue.conf.gz detail
-rw-r----- 1 root wheel 2929 Mar 24 16:12 /config/rescue.conf.gz
total 1