topic Re: Metric Data using REST API - Faulty frequency in older timestamps in Splunk AppDynamics

Metric Data using REST API - Faulty frequency in older timestamps

Ashmita_Thapar — Wed, 11 May 2022 10:21:58 GMT

I am trying to request metric data from my controller using metric-data rest api. Though, the frequency of data points is showing some inconsistent behaviour while giving time frame of current data and older dates.

e.g. Fetch Business Transactions' Calls per minute metric data for 15 minute frame in today.

Frequency - ONE_MIN in response and number of data points is 15 which is accurate.

But for older date, say 9th May 2022 data being requested on 11th May 2022 for a data frame of 15 mins, similar to above scenario. Then,

Frequency - TEN_MIN in response and only 1 data point which is inaccurate.

Note: rollup is false in both cases.

https://xx.saas.appdynamics.com/controller/rest/applications/xx/metric-data?metric-path=Service+Endpoints%7Cxx%7C%2Fxx%7CIndividual+Nodes%7Cbaseapp%7CCalls+per+Minute&time-range-type=BETWEEN_TIMES&start-time=1652107140000&end-time=1652108040000&output=json&rollup=false

[
    {
        "metricId": 2960604,
        "metricName": "BTM|Application Diagnostic Data|SEP:211315|Calls per Minute",
        "metricPath": "Service Endpoints|xx|/xx|Individual Nodes|baseapp|Calls per Minute",
        "frequency": "TEN_MIN",
        "metricValues": [
            {
                "startTimeInMillis": 1652107200000,
                "occurrences": 1,
                "current": 141,
                "min": 0,
                "max": 167,
                "useRange": true,
                "count": 10,
                "sum": 1264,
                "value": 126,
                "standardDeviation": 0
            }
        ]
    }
]

Urgent help required over this one...

Re: Metric Data using REST API - Faulty frequency in older timestamps

Ashmita_Thapar — Wed, 11 May 2022 10:55:25 GMT

Adding to my set of observations, in older dates if i request data for timeframe less than ten minutes long e.g. 5 minutes. Then, the data set is empty and no data points are returned.

Re: Metric Data using REST API - Faulty frequency in older timestamps

Morelz — Wed, 11 May 2022 12:48:18 GMT

Hi Ashmita

What you are describing is exactly how it works.

Raw per minute data is no kept indefinitely , it is aggregated at certain time periods and stored in the aggregated time sections.

You can read more here which explains how that works

https://docs.appdynamics.com/22.5/en/appdynamics-essentials/metrics-and-graphs/metric-data-resolution-over-time

Ciao

Re: Metric Data using REST API - Faulty frequency in older timestamps

Ashmita_Thapar — Mon, 16 May 2022 04:19:41 GMT

Hey Mario,

Thanks for answering my doubts. Even though i understand the concept, but i would like to understand the difference between behaviour of UI and Rest. The UI is still capable to plotting the data with frequency ONE_MIN whereas the same time frame is served with minimum frequency TEN_MIN for rest.

Also, the minimum TEN_MIN frequency restricts user from accessing data if the time frame of interest is smaller than 10 minutes.

Re: Metric Data using REST API - Faulty frequency in older timestamps

kalyan — Fri, 20 Jun 2025 13:40:58 GMT

Can i please which api you were using to get these data values?