Prometheus remote read and Influx retention policies.

[Sohaib Omar]

Hi all, 
I have InfluxDB set up for remote write/read storage for Prometheus. I lack a bit of an understanding of how does Prometheus decides which retention policy to use? Or does it even use one?

Let's say below is my remote read config

apiVersion: monitoring.coreos.com/v1

kind: Prometheus

...

remoteRead:

  - readRecent: false

    remoteTimeout: 1m

    url: http://data-influxdb.tick:<port>/api/v1/prom/read?db=monitoringdb&rp=rp_24h //rp_24 = retention policy last 24 haour

  - readRecent: false

    remoteTimeout: 1m

    url: http://data-influxdb.tick:<port>/api/v1/prom/read?db=monitoringdb&rp=rp_7d // retention policy of last 7 day, this is downsampeled every 5 minutes using Influx continuous queries.

...

say that a metric_x is present in both retention policies shard and if I query Prometheus via HTTP using below

/api/query_range?start=now&end=now-2day&step=20minutes

will Prometheus be able to find and use the right retention policy(rp_7d in the above case) or will it fail?

Thanks 

P.S: In my case, Prometheus API does not return data when time range is more than 24 hours. I couldn't find any reading about and that's why I am trying to understand the expected behavior.