[Orca-dev] Re[2]: RE[2]: Digest 336

[Rusty Carruth]

Blair Zajac <blair at orcaware.com> wrote:
> Rusty,
> 
> Since there are already 30+ plots and instead of creating 8 new plots, can
> we create one new plot that plots all of these?  That way somebody can just
> look at this one plot and see which subsystems need help?
> 
> Hopefully, most of the systems will be in a state with value 0, so the states
> that are high should stand out.
> 
> Also, the plot for
> 
> 	DNS Warning/error states (dw)
> 
> should be for DNLC (Directory Name Lookup Cache), not DNS and dw is the old
> name.

Here's my first cut at that:

plot {
title			%g Warning/error states
source			orcallator
data			State_D
data			State_N
data			State_n
data			State_s
data			State_r
data			State_k
data			State_c
data			State_m
data			State_d
data			State_i
data			State_t
line_type		area
line_type		line1
legend			Disk state 
legend			Network state 
legend			RPC Client state 
legend			Swap state 
legend			RAM state 
legend			Kerenel Memory state 
legend			CPU load state
legend			Mutex state
legend			DNLC state
legend			Inode state
legend			TCP/IP stack state
y_legend		Warning level
data_min		0
plot_min		0
}

I've not waited for the output yet, so there may be serious problems
with my cut and paste exercise above!

(I debated removing the 'state' from the legend - we'll know once I get
a plot from this addition if I should or not...)

> By the way, this is a create idea.

Thanks!  Now lets hope I can use this info to get our machines upgraded
to where we really need them to be!  (Ain't it funny how real life makes
you want to do the right thing?  ;-)

rc