Internal documentation only

This page has been marked as draft.

Can I use Geneos to estimate when a disk might fill up?

The short answer is ’not reliably’, since the use of a disk is going to change dramatically depending on the context. Edu Morales from the ITRS US office has played a little with the idea and we have the following - note that we must heavily caveat that this is not a formal solution, it’s just some experimentation.

Having played with rule periods over the last few months I tried to see if I could do something interesting with those. My first attempt was to use a history period and to get the min and max/current of this and calculate rate. While this seemed to work at first I quickly realized this would only work if all of this was ever increasing. I then thought that I could use the first value of a history period and the current value and use that for my rate.

Unfortunately the first function only works with Xpaths and not history periods. Had a quick chat with Engineering and we came up with the following. Use two history periods and get the average of each corresponding history period. This should create a reasonable approximation of the growth as long as the data set is reasonably linear.

The formula is then:

rate =  ( avg(X) - avg(2X) ) /  (X/2)

It looks something like:

set $(first) average(value for "Rolling10Minute")
set $(second) average(value for "Rolling5Minute")
set $(rate) ($(second) - $(first)) / 150
if $(rate) > 0 then
 set $(secondsTilFailure) (100 - value) / $(rate)
else
 set $(secondsTilFailure) "undef"
endif

Example of semi-linear data and rate function using Microsoft Excel compared to Geneos Rate.

Note

History periods require gateways to retain these points in memory, so I would recommend not going overboard with this. Smaller history periods have less data.