The system log is in /var/log/system. It can grow large, and being on an unstable system, I'd like to recreate the error on a smaller log file before posting it. Is there a way for me to manually initiate the use of a new system log file?

Thanks,

Do you have logrotate installed? It should help manage the size of the log files.

If this is an error you can reproduce at will, you could run

and capture the output to the console as the error is logged.

Edit:

I just learned about this command in this thread. It may do what you want.

Either use the "tail -f" as suggested above, or ...

mv -f /var/log/system /tmp/old_system_log; killall -HUP <syslog_process>

Hope this helps,

/var/run/syslogd.pid contains the process id of syslogd. Every month I kill that process, rename /var/log/syslog and the other log files to which syslogd writes, adding the year and month, e.g. syslog becomes syslog-201412, then start syslogd again. On my distribution I can use /etc/rc.d/rc.syslog to start and stop syslogd, but that's unnecessary.

I reserve syslog for the most important system information by separating non-essential stuff into separate log files, most importantly the dhcp stuff (and lynx, for me, but who else uses that anymore?). I did the same for acpi and apm back when I had them.