[rrd-developers] implementing portable format

[Tobias Oetiker]

Thursday kevin brintnall wrote:
> > when thinking about the new format, it might also be worth spending
> > some thought on shortcomings of the present format. I think it
> > might be possible to add the following features quite easliy
> >
> > * unlimited string length for labels
> > With unlimited I mean that it is parametrized and gets set at rrd
> > creation time.
>
> Have there been many cases where the label length is not sufficient?

every now and then there are users complaining about this ... and
it is sort of embarassing since there is no real reason to have
this length fixed as fahr as I can see. It would have to gro in 8
byte steps, but thats about it ...

> > * ability to store 'user data' in the rrd header and along with each
> > datasource
>
> For example?

some people like to store meta data at the rrd or ds level.
obviously this can be done easily in a second file sitting next to
the rrd, but then again as we change the format something like this
could be added fairly easily.

> > Also the sizes on scratch space in PDP and CPD areas could be
> > paramtetrized which would help with future enhancements of the data
> > format.
>
> Why would they need to be variable size?  I'm a little fuzzy on the
> inner-workings of [pc]dp_prep.

if you look at the implementation of the COMPUTE datasource type,
then you will see that there were various tricks employed to make
do with the available space ... I asume there could be other
datasource types or consolidation methods that require additional
scratch space for calculation. by providing the ability in the
format, such types could be added without any further format
changes. appart from the handling of the new DS type or
consolidation method.

cheers
tobi


>

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
http://it.oetiker.ch tobi at oetiker.ch ++41 62 775 9902 / sb: -9900