List users,
Please provide me with tips on how to replicate a single file to a
separate machine as changes are made to it. I would prefer a method
that reacts to file modifications (ie. FAM/gamin) as opposed to timed
loops/polling (cron + rsync). I'd also like to avoid NFS altogether.
Keeping resource consumption low on the source machine is a priority. A
bit of research has lead me to believe that calling rsync when gamin is
alerted to a file modification would be a good fit for my scenario, but
I'm unclear on the easiest implementation.
My scenario is as follows. I have a machine that runs Asterisk VoIP PBX
software. Asterisk creates a log file that we generate reports off of.
Another machine handles the generation of these reports, which involves
significant number crunching and file I/O. By replicating the file on
the reporting machine, I'd like to decouple the resource consumption of
reporting from the VoIP server. Some of the reports are used to monitor
activities in realtime, so cronning off rsync on a large time interval
is not an option.
Thank you,
Matthew Roth
InterMedia Marketing Solutions
Software Engineer and Systems Developer
