[Orca-users] Orca messes up on missing data?

[Danny Howard]

Hello,

I have a data file like so:

load1 load5 load15 disk-/ disk-/home disk-/tmp disk-/usr disk-/var 
numproc-total numproc-httpd numproc-httpsd numproc-mysqld qmail-success 
qmail-deferral qmail-failure connect-imapd connect-pop3d mem_free 
mem_wire mem_active mem_inactive mem_cache
0.22 1.07 0.83 2 15 1 8 0 92 19 9 1 2978 782 515 45 4808 1377513472 
287318016 272515072 153317376 7790592

Orca seems to assume that data will always appear in the exact same 
solumn.  If ever the ordering of columns changes, I'll see aberrant 
behavior, like data spikes, in the Orca graphs.  This is especially 
annoying when I change my collection scripts, as i have to restart 
Orca.  My collection scripts will also sometimes omit data that they 
were unable to collect, and I'll see, say, qmail deliveries on a host 
spike from 5/minute to 6,000/minute, which pretty much ruins the graphs.

Any suggestions on a workaround?  I dimly recall running into this in a 
past life, and I resorted to killing Orca's state file and running it 
"once" from cron.  If I could just get Orca to sanity-check that the 
numbers in the columns on the second line are in fact tied with the 
labels on the first line ... or is something ele going on?

Sincerely,
-danny

-- 
Danny Howard                                    djh at servercentral.net
Technical Support Manager                          (312)829-1111 x235
Server Central Network                   http://www.servercentral.net