[rrd-users] Re: Who has the most RRD files (or data sources)?
jfesler at gigo.com
Mon Mar 20 23:30:41 MET 2006
> Does this boil down to meaning that the parts of a single RRD file are
> physically close to each other on disk? And that multiple RRD files are
> further apart? I can see how that would make things a bit faster or slower
> but it doesn't seem like it would make the huge difference that some people
> seem to see when they put multiple RRAs into one RRD.
Whatever one works, please use it, and look for seek/read/write calls.
> Am I just clueless? :) Or does it make sense to you that the difference is
Yes, and yes.
Each RRD update involves seek/read/seek/write the header, and seek/write
once per RRA. Doesn't matter (realistically) how many DS's there are for
rrd_update.c is the authoritice source on what really happens under the
hood, but strace/truss and the like can also give you a view of what is
happening - without having to know C.
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help mailto:rrd-users-request at list.ee.ethz.ch?subject=help
More information about the rrd-users