[rrd-users] Time Shift on Trend Graph

[Chris Mason]

Hi Karl

while this is not cheap, it's fairly easy. I've used it a

> couple of times since I needed the first valid entry as well.
> It takes a few seconds - depending on the size of the rrd -
> and spits out the timestamp of the first non-NaN entry.
>
>
> cd /tmp
> rrdtool dump yourfile.rrd | csplit - '%<database>%+1' '/<database>/+1'
> '{*}'
> grep -Ehivm 1 '<row>(<v> *NaN *</v>)+</row>' xx?? | sed 's:^.* /
> \([0-9]\{10\}\) .*:\1:' | sort -n | head -1
> rm -f xx??
>

Perfect, Thank you!

However, I need to impact the performance implication on doing this across
multiple RRDs.
Anyone know, how likely the format of a "dump" is going to change between
different versions of rrdtool?

Thanks,
Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.oetiker.ch/pipermail/rrd-users/attachments/20130702/cb66be7b/attachment.htm