Just a quick one: I’ve setup the DNS latency collector as per:
The chart for latency looks OK but the one for query status shows all as returning a value of ‘0’. This seems count-intuitive to me; is there any good reason why it’s a 0 instead of a 1?
Hi, @Luis_Johnstone. The query_status chart has 3 dimensions: success, network_error, dns_error.
- Each of them represents a state.
- The values can be 0 or 1.
- 0 means not in the state.
- 1 means in the state.
Only one of them can be 1 (active) at a time (read as “the current state is …”).
Ah, I see; it’s because you need any of the other 3 dimensions of the chart to also be able to be a 1 to indicate a problem (e.g. it wouldn’t make sense if network_error = 1 meant everything was OK). I figured that it was something like that. Thanks!!
It is like this @Luis_Johnstonebecause because aggregation over time is meaningful.
For instance, you get an average value for every state for the last 30 minutes and they are:
- success: 0.7
- network_error: 0.2
- dns_error: 0.1
This means that in the last 30 minutes, the percentage of successful requests is 70%, and so on.