Anomaly detection with timeframe

[audiokarate]

Hi, im new to dynatrace and Im having trouble creating an anomaly detection that will verify the increase of a metric for a period of 24h. 

Here is the query im running in the data explorer page.

However when I create my anomaly detection it doesnt include the 24 hours filter... Do i need to modify my query to add a timeframe if so how can i do this?

[PacoPorro]

Maybe you are looking for this the https://docs.dynatrace.com/docs/shortlink/api-metrics-v2-selector#timeshift ?

[audiokarate]

i tried that before opening the ticket and its not giving me the result i need

i need to be alerted if my metric does not increase by at least 1 in the last 24 hours...

see screenshot below

[PacoPorro]

Try to do a calculation

(your metrics with timeshift) - (metric without timeshift)

[audiokarate]

im having the same problem

[PacoPorro]

So, livraison and observability-cac ran 163 successful backup in the last 24h, and also 163 in the interval from -42h to -18h. 
Looks like the expected behaviour is to have the same number of backups in the last 24h than in previous period.

[audiokarate]

actually we run only 4 backups a day (every 6h) and i have to make sure that there was atleast one sucessful backup in the last 24h.

I have no idea where this 163 value is coming from.

We are in the process of doing the migration of our monitoring from prometheus to dynatrace. If i compare the exact same metric in dynatrace and in in prometheus im getting different results (see both graphs below)

[PacoPorro]

It looks like you are using fold operator.