2007-11-20 #2

Useful:

flowdb=# select count(ip) from localip;
 count
--------
 424291
(1 row)
flowdb=# select count(distinct ip) from localip;
 count
-------
   111
(1 row)
flowdb=# select count(distinct timeslot) from localip;
 count
-------
  3153
(1 row)

Queries 1 and 3 were pretty fast; query 2 took about 15 seconds.

Data import took 292 minutes, but I have to run it again after adding the backup VLAN address definitions.

Added the --reset flag to postgres2rrd. I basically wrapped my "template" perl script with all it's default functions around the existing script. Just goes to show you should never try to cut corners, you'll end up having to do it correctly eventually. I still have to do this for the other scripts, although graph is just a place holder (but I've said this before about the other ones as well)...

Changed the duplicate-import detection in flow2postgres; previously it saved the path/name as invoked, which changes depending on how we import all the files. So now it just saves the name (not the path/name) which should make import-duplication detection easier/correct.

More/different ToDo:

• wrap the flow2posgres in the perl template
• make the output for all scripts either 'debug' or 'verbose' output
• maybe build poor-man's timing information into the scripts?
• consolidate the ToDo lists into one list.
• the duplicate-load detection in posgres2rrd isn't working again.