Open abelfodil opened 9 months ago
I second this.
I have a multigig egress and I noticed I suffer from fluctuation quite a bit (e.g., when a known-slow server was automatically selected).
I don't particularly care about a one-off speed dip, it is only when it is consistently slow that I'd like to be notified.
Echo'ing @abelfodil's comment in terms of implementation it should be of a moving window, where user can set M out of N
as a notification threshold.
When M
of the most recent N
results are breaching, fire off the notification.
Is your feature request related to a problem? Please describe.
Often times (~25%) my speedtest retuns low results for 1 or 2 data points that breach the thresholds but then recovers. This leads to false positives or at least noisy behavior.
Describe the solution you'd like
I would to have the possibility to specify a number of data points that are being breached before notifying me that there's an issue.
For example, if I put 3 as the "Datapoints to Notify", then threshold must be breached 3 consecutive times before notifying me.
Additional context
For example, see
Datapoints to Alarm
in https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/AlarmThatSendsEmail.html#alarm-evaluation