[rrd-users] Using a single RRD vs. a separate file for each data series

[seanadams]

Sorry for the re-post but I don't think this went through to the mail gateway
the first time. I'm not really looking for help with my particular
application, but rather I would like to understand this issue in a general
sense: 

It appears that one of the first decisions an rrdtool user must make is
whether to store all their data in a single RRD file, or use a separate RRD
file for each series (or perhaps some hybrid scheme). 

So, what exactly are the implications of have a single RRD containing
multiple series, versus multiple RRD files?

In my application I expect to add a new data series occasionally, so it
seems like I should have a separate RRD or each value I want to track. It
appears that rrdgraph can easily read from multiple files, so I don't see a
problem. 

I understand that using a single RRD requires all values to be written
simultaneously, but I don't see any obvious advantage to having a single
RRD. Is it done to improve CPU or storage efficiency? 

Are there any problems that I might run into if I use separate RRDs - for
example, operations involving multiple series that I won't be able to do? 

thanks, 
Sean
-- 
View this message in context: http://www.nabble.com/Using-a-single-RRD-vs.-a-separate-file-for-each-data-series-tp19439775p19439775.html
Sent from the RRDTool - User mailing list archive at Nabble.com.