Sun Enterprise 10000 Server
 

PLA-Support>Sun Systems>SPARC>Enterprise>SN-SPARC: SF-Exxk
 
.Old GCS Categories>Sun Microsystems>Servers>High-End Servers
 

Applies to:

Symptoms

Sometimes when attempting to start a netcon session to a Starfire[TM] domain, you just get a "trying to connect" message without seeing "connected." even if you wait several minutes.

Example:

starfire-ssp:Starfire1% netcon
trying to connect...

You may find that you can telnet to the domain ok, that the OS is working without problems, and that the cvcd is running. 

Cause

If this is the case, then the most common cause is that the netcon_server process for that domain which runs on the SSP has died.

Solution

netcon_server is started by bringup and should remain running as long as there is either an OBP or an OS on the domain.
During an SSP reboot it is automatically restarted by startup.tcl, but other than that it is not monitored by the startup.tcl script.

If you want to be sure which netcon_server process relates to which domain (there should only be one per domain), you can use the following command:

/usr/ucb/ps -axwwe | grep netcon_server | grep -v grep

(NOTE: The output of this command is extremely verbose, as you get about 16 lines per process, and you can't reliably use grep to shorten it further)
Check the value of SUNW_HOSTNAME which is displayed for each netcon_server process.

To restart the netcon_server for a domain, use:

domain_switch <domainname>

netcon_server -r

This is the same way startup.tcl *reconnects* to netcon_server to the domain.

Product
Sun Enterprise 10000 Server

Keywords: netcon, netcon_server, Starfire[TM], ssp

Internal Section

Previously Published As 16764

Also check Document 1005993.1 (previously published as SRDB 70018) for further info about the issue.

Product_uuid: 29ddef46-0a18-11d6-92de-ae47474f0f6c Sun Enterprise 10000 Server

This solution has no attachment