[Orca-dev] Errors with procallator

[rhfreeman at micron.com]

Hi Guilherme,

I was hoping you'd be on this list! :-)

Ahem, yes, df -l -k does make far more sense and I've changed our local version to use that instead. Monitoring the remote space on all our machine doesn't make much sense, especially seeing as all the /home directories are automounted and the mount list could quite long and it is lot of duplicate/pointless effort.

If you've got a new version I could use, then I'd be happy to test it for you!

Here is the bounce message I got, I used the email address form the README:

Hi. This is the qmail-send program at dominios.abordo.com.br.
I'm afraid I wasn't able to deliver your message to the following addresses.
This is a permanent error; I've given up. Sorry it didn't work out.
<gchehab at abordo.com.br>:
user does not exist, but will deliver to /dev/null/Maildir/
can not open new email file errno=20 file=/dev/null/Maildir/tmp/1057328841.11290.dominios,S=2755
system error

Cheers,

Rich

-----Original Message-----
From: Guilherme Carvalho Chehab [mailto:Guilherme.Chehab at tco.net.br] 
Sent: Friday, July 04, 2003 1:59 PM
To: 'orca-dev at orcaware.com'
Subject: RES: [Orca-dev] Errors with procallator


Hi Rich, 
One better approach is to change the "df -k" for "df -l k", that asks "df" to return only local mounts, excluding all remote FS (including NFS). As monitor their space usage in the clients if a doubled effort (as they are probably monitored on the hosts already).
You said that your email to me bounced, that´s strange, what e-mail address did you tried ? 
I am really not giving to much attention on maintainnig procallator lately, bue it is quite stable and have a minor number of glitches, I am working on some patches on my spare time (which has been quite few :-)
regards, 
Guilherme 


> -----Mensagem original----- 
> De: rhfreeman at micron.com [mailto:rhfreeman at micron.com] 
> Enviada em: sexta-feira, 4 de julho de 2003 06:09 
> Para: orca-dev at orcaware.com 
> Assunto: [Orca-dev] Errors with procallator 
> 
> 
> Hi folks, 
> 
> I did try and email the author of procallator about this, but 
> the email bounced. So I thought I'd send it here instead. 
> 
> This is procallator from orca-0.27, and it is running on RH 7.2. 
> 
> When running it, my machines were giving this error: 
> 
> Jul  3 15:25:00 avalanche procallator.pl: Use of 
> uninitialized value in numeric eq (=3D=3D) at 
> /tools/Orcaware/orca/bin/procallator.pl line 499. Jul  3 
> 15:25:00 avalanche procallator.pl: Use of uninitialized value 
> in concatenation (.) or string at 
> /tools/Orcaware/orca/bin/procallator.pl 
> line 220. 
> 
> I tracked it down to the "df -k" statement. Some of our 
> output is below: 
> 
> /dev/hda3              4127108    247164   3670296   7% /var 
> MrPotatoHead:/usr/local/lsf 
>                      297317840 221648608  75669232  75% /usr/local/lsf 
> 
> The problem is that if the mount name is long, then it line 
> wraps and causes problems. The mounts with long names are NFS 
> mounts, and we don't want these monitored anyway. So I 
> changed a line to only do this for lines beginning with /dev, 
> which means locally mounted disks only: 
> 
>       if (!(($df[$i]=3D~/cdrom/)||($df[$i]=3D~/cdrom/)) && 
> ($df[$i]=3D~/^\/dev/)) { 
> 
> I'm not sure why cdrom is in there twice, but I left it alone. 
> 
> Also, for the start-up procallator script I added this line: 
> 
> export COLUMNS=3D200 
> 
> Otherwise ps doesn't return a line long enough to detect procallator. 
> 
> If you've got an updated version, then let me know and I'll 
> give that a try. It looks like there is a newer version in subversion: 
> 
http://svn.orcaware.com:8000/repos/trunk/orca/data_gatherers/procallator 
/ 
Shall I just use that version instead? 
Cheers, 
Rich 
_______________________________________________ 
Orca-dev mailing list 
Orca-dev at orcaware.com http://www.orcaware.com/mailman/listinfo/orca-dev