[rrd-developers] DS idea

[Alex van den Bogaerdt]

----- Original Message ----- 
From: "Steve Shipway" <s.shipway at auckland.ac.nz>
To: "Alex van den Bogaerdt" <alex at vandenbogaerdt.nl>; 
<rrd-developers at lists.oetiker.ch>
Sent: Monday, October 11, 2010 2:05 AM
Subject: RE: [rrd-developers] DS idea


>> >>   GAUGELATEST
>> >
>> > Whatever it's called, I'd certainly like to see a DS type which is
>> like
>> > GAUGE but without the normalisation to the interval.  There are many
>> > applications for which this would be useful.
>> >
>>
>> I'm probably missing something here.
>>
>> What can you do with that, which cannot be achieved with step==1,
>> steps=={whatever},heartbeat>={whatever} and CF LAST (or, as in this
>> thread, MAX) ?
>
> If this new type is created, then you don't need to create an RRA with a 
> step of 1, it will work however long the step interval is.

There is no need to have an RRA with a step of 1. Maybe you mean RRD?

> This means that all the standard RRA definitions you're using (many people 
> have a step of 5min) will work without change.  In particular, MRTG will 
> be able to use this new data type with minimal alteration, as will other 
> software that uses the RRD API.
>
> While you can manage to approximate this by special configurations of DS 
> and xff setting in the RRA, it is much simpler if you can simply set a 
> data type.  I think this would be a good feature to support directly, 
> rather than via more complex settings...

I'm sorry, but you have not convinced me. There is nothing special about 
that configuration, except that the numbers to use are different than MRTG 
uses. Creating a new DST, just to avoid having to alter some parameters, 
still sounds wrong. Now *that* would be special and complex. And it 
introduces more code in which bugs can hide.

And where does the idea of having to alter xff from?  I can see why a newbie 
could think this, but you ?!?

I currently have no access to an rrdtool installation to verify the 
following, so I rely on others to tell me which part of the following does 
not work:

MRTG-alike settings, with an XFF of zero:

rrdtool create name.rrd --step 300 \
DS:ds0:GAUGE:600:U:U \
DS:ds1:GAUGE:600:U:U \
RRA:AVERAGE:0:1:600 \
RRA:AVERAGE:0:6:700 \
RRA:AVERAGE:0:24:775 \
RRA:AVERAGE:0:288:797 \
RRA:MAX:0:1:600 \
RRA:MAX:0:6:700 \
RRA:MAX:0:24:775 \
RRA:MAX:0:288:797

Similar settings, but keeping track of the "real" MAX, and XFF still being 
zero:

rrdtool create name.rrd --step 1 \
DS:ds0:GAUGE:600:U:U \
DS:ds1:GAUGE:600:U:U \
RRA:AVERAGE:0:300:600 \
RRA:AVERAGE:0:1800:700 \
RRA:AVERAGE:0:7200:775 \
RRA:AVERAGE:0:86400:797 \
RRA:MAX:0:300:600 \
RRA:MAX:0:1800:700 \
RRA:MAX:0:7200:775 \
RRA:MAX:0:86400:797

I did nothing more than change step from 300 into 1, and multiplying each 
steps by 300. Nothing special, nothing complex, just different.

And I'm sure how you've noticed that I can use COUNTER instead of GAUGE. No 
need to program "SPECIALGAUGE", "SPECIALCOUNTER" and so on.