[Orca-users] Multiple disk space percentage graphs

[Tonij T.]

Thanks for the information, Dragon .  I realized that when I replied I didnt 
CC: the mailing list, and since the point of mailing lists is to share 
information I thought I would update everyone on this.

He did a great job of pointing me in the right direction.  I have been 
fiddling with the orcallator.cfg file to see how I can change things.  For 
some reason, no matter what changes I make to this file, the graphs never 
change.  Not sure what is going on there but all is not lost because I have 
another idea anyway...(I know, I am all over the place here but bear with 
me... :D )

Since my main goal here is to isolate specific file systems, why not go 
right to the source?  Since Orca  is using RRD to collect disk space metrics 
I decided to learn how it is being stored and how Orca is pulling that data 
out.  What started out as a daunting task has actually yielded some decent 
results.  I am able to create manual graphs of specifc file systems on the 
fly now.  Cool! I love RRD, its an awesome tool.

Now, what would be a BIG help, is if I could see just how Orca is generating 
these graphs.  The nuts and bolts of the .png creation, I guess.  I have the 
basics down, i.e. I can create a graph with a given start time and end time 
using the counter for the given filesystem.   What I am a bit baffled on is 
how Orca is also showing me averages, min and max values.  I have been 
trying to get that into my custom graphs but have not had any luck.

Where does Orca do all the RRD work?  I dont see anything in orcallator.cfg 
that is actually creating the graphs but I do see some variables.  It would 
be a huge help if I could see something like:

rrdtool graph image.png  -u 100 -l 0 -v "Percent Used"  --start 1138600800 
--end 1139501009   
DEF:used=gauge_mntP_per_SOME_FILE_SYSTEM.rrd:Orca19990222:AVERAGE  
LINE1:used#FF0000 -variable -variable -variable etc.

Am I making sense to anyone? :D

Oh and another question:  I am using Orca version 027.  I dont see a later 
version anywhere but I did find something called "orca-snapshot-r525".  Is 
that an update?  Its all rather confusing to me, I would expect a newer 
version to be 028 or 030 or something...



>From: David Michaels <dragon at raytheon.com>
>To: "Tonij T." <tonij67 at hotmail.com>, orca-users at orcaware.com
>Subject: Re: [Orca-users] Multiple disk space percentage graphs
>Date: Wed, 08 Feb 2006 11:29:16 -0700
>
>Tonij --
>
>You can specify the plot with with the "plot_width" setting in the plot {}. 
>  For instance:
>
>plot {
>title                   %g Disk Inode Percent Usage
>source                  orcallator
>data                    mntp_(/(?!vxckpt|checkp|verita).*)
>...
>plot_width      700
>}
>
>As for having too many filesystems on one graph, you can create multiple 
>plot {} entries, and just modify the data expression to include a subset of 
>those filesystems.  For example, if you have lots of filesystems under 
>/export, you could do something to specify /export/[a-m]* for one plot and 
>/export/[n-z]* for another.  In my example above, I exclude any veritas 
>checkpoint filesystems.
>
>Also, if you're not sure which 'blue' you're looking at, you can compare 
>the values printed for those blue filesystems in the bottom of the graph.  
>Usually the max / current numbers will help you identify which line is 
>which.
>
>Lastly, you can specify a color order using "color", as in the above 
>example.
>
>Cheers,
>--Dragon
>
>
>
>Tonij T. wrote:
>>Hi all,
>>
>>I have been experimenting with Orca on and off over the past year and have 
>>finally gotten some good trending information over the past few months.  
>>One of the main metrics I am interested in is trending disk space usage.   
>>I have two questions about this feature:
>>
>>1)  I am monitoring 12 Sun systems.  For each system on the "Disk Space 
>>Percent Usage" page there can be up to 50 (!) links to see the graph.  
>>Some of the larger systems with upwards of 60+ file systems can take up to 
>>5 minutes to populate this page.  Is there a reason this graph behaves 
>>this way?
>>
>>2) Speaking of the larger systems, when I try to look at the graphs for 
>>some of these, there are just so many of them the graph is a clutered 
>>mess.  With 10 different shades of blue I really can't tell which file 
>>system I am looking at.  Is there a way to resize this graph to make it 
>>more readable?
>>
>>Thanks in advance!
>>
>>
>>_______________________________________________
>>Orca-users mailing list
>>Orca-users at orcaware.com
>>http://www.orcaware.com/mailman/listinfo/orca-users
>