Currently, we can do checks down to a minute for websites in the app; but you can only get down to hourly data from the API. It would be nice to have the option to get this minute granular data via the API calls, maybe adding an extra arg in the call could work.
Current use:
'?period=0&unit_of_time=2'
Posible use:
'?type=URL&period=0&unit_of_time=2'
/api/reports/availability_summary/{monitor_id}?period={report_period_constants} doc
/api/reports/availability_summary/type/{monitor_type}?period={report_period_constants} doc
Hope you are talking about the these availability summary report apis.
Availability, which is a percentage value, is calculated for hourly intervals. If we return availability data per minute, for a monitor having the minimum one minute polling interval, the returned value of availability for each minute will be either 100% or 0% except for the minute when the status change has occurred. This becomes more of a heat map than availability percentage chart.
If the heat map is your requirement, you can get the duration for outages and consider monitor is available for the rest of the time. The same api contains "outage_details" node which will give your the exact duration of each outage occurred during the given period.
Let me know if your requirement is different than this.
Basically what we are doing is grabbing the data and entering it into a database that is used to generate a specific report from our direct customers. We currently are pulling data but more granularity came up as a possible need n the future.
If I understand you correctly, if we drop the granularity to a minute for the current hour then we could technically get the actual minute data from the database and then create our own calculations against that correct? Or would the data be skewed as it would only be looking at the current poll which would only tell us that it's either up or down?
I think I understand it and would say it would not be useful since we'd have to do the calculations you are already doing if we get hourly.
Please correct me if I'm wrong, thanks.
Yes, if you want more granular information, say availability percentage for every 15 minutes window, you would end up doing the calculation that we are doing to arrive the percentage value.
So for the availability data, unless it is a pressing need, I would recommend to continue with the hourly percentage value that we provide.
For the performance data, we are already providing value for every poll made during the given period by passing unit_of_time (granularity) as 1 (Raw Data).