Order Tray | Contact Us | Home | SIG Lists

[nos-bbs] Jnos2 interface trace to file

Michael Fox - N6MEF n6mef at mefox.org
Fri May 6 18:30:34 UTC 2011


In the current implementation of JNOS2.0, you can specify a file name and a
file sync interval (in milliseconds) as additional parameters to the trace
command.  

On our systems, we typically have 3 interfaces running and all are logging
to a trace file at 1/2 sync intervals.  

I like the faster sync interval because if I need to see something in real
time, I can just "tail -f" the trace file and watch what's going on.  If
you're not interested in real time, then a one or two second sync might use
less resources on your machine.  We've tested with all three interfaces
running at 1/10 sec sync interval with no problems.  But we keep it at 500
msec.  

I don't have the manual in front of me.  But just put the trace command with
the file name in your autoexec and add a sync interval, (such as 1000 for 1
second) at the end of the command.  We use logrotate to rotate the trace
files and we maintain 30 days of traces for each interface.  Typically, we
don't need to go back more than a week.  But occasionally someone will say,
"you know, when I checked into the Tuesday net a couple of weeks ago this
weird thing happened ..."

Michael


-----Original Message-----
From: nos-bbs-bounces at tapr.org [mailto:nos-bbs-bounces at tapr.org] On Behalf
Of Joe Erlewein
Sent: Friday, May 06, 2011 10:25 AM
To: TAPR xNOS Mailing List
Subject: [nos-bbs] Jnos2 interface trace to file

All,
In my experience so far running a headless virtualized Jnos2 server, I've
made a looooong list of things to check.
(My ultimate goal is a small footprint transportable basic config under
Virtualbox, physical interfaces connected to Digi PortServers over ethernet,
connecting to TNCs, etc)

back to the point of this email.. 
"on-air" traffic (trace vhf 0111) that I'm capturing in 'screen' (manually
using c-A H) is minimal... compared to standard logging to dmesg/syslog
resources.

I'd like to (at least temporarily) log tracing on all interfaces to logfiles
for later analysis.. somehow "other", i mean, than manually telnetting to
localhost, logging in, @ command, turning on tracing, and then starting a
screenlog capture.

Any ideas?

Thanks!
-jre
_______________________________________________
nos-bbs mailing list
nos-bbs at tapr.org
https://www.tapr.org/cgi-bin/mailman/listinfo/nos-bbs




More information about the nos-bbs mailing list