Cheshire Cat Computing

This is due to the delay between you viewing the data, and the data actually being collected.

eg,

if you collect data at 12:00, 12:05, 12:10.... and view at 12:04

then the graph will show 'unknown' for the last 4 minutes (since it doesnt know the state after 12:00).

There are a couple of ways around this. First, you can set the 'uselastupdate = yes' optionin the routers2.conf file, which will always end the graph at the last update time (usually the best solution).

Secondly, the code makes a few allowances for this and winds back one data point under some circumstances. This does not always work, and it has no way to tell the difference between something that is really down or that is just late in its data collection.

So - the graph is (technically) correct, although it is not showing what is preferred. I'm going to look into improving the code that avoids this grey bar.

Steve

with regard to the 100% available problem -- on our system it shows 99.75% -- and I agree a neat 100% would look better. This hits the same problem as with the last grey bar, though, in that the system cant work out for sure if it has just gone down, or if it is just at the end of the data collection window. Maybe I should use 'current time - 5min' as the base? Although, this would cut out the last data point if we had updated this minute....

Steve