# Huge Difference in Graphs of Prometheus & InfluxDB for same data **URL:** \ **Category:** Systems\ **Tags:** influxdb, prometheus, grafana, influxql\ **Created:** [June 20, 2020, 7:42am UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912 "2020-06-20T07:42:19Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![yogi](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/yogi/32/4912_2.png) [@yogi](https://community.influxdata.com/u/yogi)\ **Post date:** [June 20, 2020, 7:42am UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/1 "2020-06-20T07:42:19Z")
I am collecting data of a network device over SNMP & storing it in Prometheus & InfluxDB ( via Prometheus Remote write). But the graph’s ranges are drastically different, Prometheus in Mbps range while Influx in Gbps range. Why? ![Screenshot from 2020-06-20 12-57-02](https://us1.discourse-cdn.com/flex023/uploads/influxdata/original/2X/e/e7d68202095fedf2e8b2db3a89b8f714cc18021a.png) ![Screenshot from 2020-06-20 12-56-57](https://us1.discourse-cdn.com/flex023/uploads/influxdata/original/2X/3/3b634f0d75dfacb7660b0e6146e9582242ea33da.png) ---
**Author:** ![yogi](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/yogi/32/4912_2.png) [@yogi](https://community.influxdata.com/u/yogi)\ **Post date:** [June 20, 2020, 7:43am UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/2 "2020-06-20T07:43:56Z")
I am scraping the data at 15 secons interval ---
**Author:** ![Anaisdg](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/anaisdg/32/6401_2.png) [@Anaisdg](https://community.influxdata.com/u/Anaisdg)\ **Post date:** [June 24, 2020, 6:45pm UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/3 "2020-06-24T18:45:08Z")
Hello @yogi, How are you collecting writing the data to InfluxDB and Prometheus? Do you know which unit you expect the data to be returned in for your use case? Thank you. ---
**Author:** ![Mert](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/mert/32/5152_2.png) [@Mert](https://community.influxdata.com/u/Mert)\ **Post date:** [June 24, 2020, 9:19pm UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/4 "2020-06-24T21:19:25Z")
The queries are not the same. You should see similar result if you set group by(1m) and derivative(1s) in influxql. Right now, you are grouping by $interval in influx query which is grafana’s dynamic variable and grouping by 1m in promql. Can try and update us with the result. @yogi ---
**Author:** ![yogi](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/yogi/32/4912_2.png) [@yogi](https://community.influxdata.com/u/yogi)\ **Post date:** [June 25, 2020, 7:26am UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/5 "2020-06-25T07:26:14Z")
Thanks for the reply @Anaisdg. I am using Prometheus’s SNMP\_EXPORTER to collect the data & store it in Prometheus for short term retention. I am storing the data collected by the Prometheus into InfluxDB using Prometheus’s remote read/write API. The problem is I am not sure about the unit. When I compare it with other tools like Preseem, LibreNMS, OpenNMS, etc, the results are different (their scrape interval is also different & avg/grouping interval also changes it). The intention is to use InfluxDB as long term retention (at least a couple of years) using downsampling (reduced data granularity over time) using CQs & RPs ---
**Author:** ![yogi](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/yogi/32/4912_2.png) [@yogi](https://community.influxdata.com/u/yogi)\ **Post date:** [June 25, 2020, 7:29am UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/6 "2020-06-25T07:29:15Z")
Thank you very much @Mert ! They match now. Can you please direct/guide me towards more information about group by and derivative apart from documentation? I didn’t find much info about them. ---
**Author:** ![Mert](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/mert/32/5152_2.png) [@Mert](https://community.influxdata.com/u/Mert)\ **Post date:** [June 25, 2020, 8:15am UTC](https://community.influxdata.com/t/huge-difference-in-graphs-of-prometheus-influxdb-for-same-data/14912/7 "2020-06-25T08:15:38Z")
Here, a great talk about irate in promql from grafanacon [![](https://us1.discourse-cdn.com/flex023/uploads/influxdata/original/2X/8/863debf9ac708bf131d7db0afaaa13814c7ca80b.jpeg "GrafanaCONline: Prometheus rate queries in Grafana") ](https://www.youtube.com/watch?v=09bR9kJczKM) I don’t have such detailed documentation actually. > **[InfluxQL functions | InfluxDB OSS 1.8 Documentation](https://docs.influxdata.com/influxdb/v1.8/query_language/functions/#non-negative-derivative)** > > Aggregate, select, transform, and predict data with InfluxQL functions.