operations-portal-users

Marcin Radecki wrote:

Hi Cyril,

I'd like to ask you for doing small investigation if you find some time for a puzzle. I'm looking at January metrics for NGI_PL and something strange happened there around 16.01.2011.

on 16 there was 15 new alarms, 0 closed, 0 assigned.
on 17 there was 20 new alarms and 0 older than 24h. Shouldn't not-closed 15 alarms from the previous day fall into 24><48 category?

Could this difference come out of the fact of regional vo.plgrid.pl alarms? Are they counted into the metrics?

Also 18 Jan looks strange... 12 new alarms and 12 older than 48h and 0 in <24h. What should be the correct relation between "New alarms" and "<24h A"?

Thanks in advance,
Marcin

Hi Marcin

As you know there are 2 kinds of metrics : dynamical metrics and the ones generated once a day.
We have corrected the dynamical metrics to avoid to take into account the vo plgrid but not the metrics generated in the morning.
This explains the non coherency between the 2 days.

I will investigate to find a solution .

New alarms = all alarms not assigned
New alarms = N24 + N48 + N72

Cheers,