Alerts for high bandwidth usage should happen sooner/smarter

[lakesol]

Issue description

Hi,

Last month we had a high bandwidth usage issue on one of our sites. The 1st we knew about it was when the alert email triggered at 90% usage. When we investigated we found an issue which we quickly resolved to reduce daily usage. At the point in the month when the alert occurred, there was no way to see out the remainder of the month with the bandwidth capacity remaining.

In our case we had used 90% of the available capacity at 77% the way through the month.

Can I therefor suggest that a more granular/smarter approach is taken to these alerts. My thoughts on this is that it should be trend based, i.e. if the current daily average usage so far in the month, multiplied by the total days in that month result in a figure greater than a threshold (90%) then a alert is sent. Obviously consideration need to be given to spikes in traffic at the start of the month which could lead to false positive alerts.

I feel if this issue can be detected earlier it will give developers enough time to take action.

Thanks

[sajumb]

Thanks for the feature request. We understand the need for the notification in some cases to be more graduated than is currently the case. We will assess the possibilities and hopefully come up with a solution that will cover all scenarios and ensure that in all cases you have sufficient time to react to a sudden increase in bandwidth consumption.