I read gatling official blog and understood that we cannot get 95th,99th percentile "All aggregations will result in computing averages on percentiles and will inherently be broken". I dint understand one thing why we are not getting simple response time series in gatling which we can customized through our queries to get any percentile we want . Without correct percentile such integration are worthless.
Is there any way we can get close to the gatling reports timing in InfluxDB if not exact.
Gatling doesn't really send to influxDB the response time of each operation. If gatling sent complete information about the operations, it would significantly reduce the performance of gatling and influxDB. You would not have enough RAM for a test with a load intensity of even 100 transactions per second. Instead, gatling sends aggregated response times over the time interval you specified in gatling.conf. Because of this, grafana cannot show us the actual value of the 90th percentile. There are many utilities for proper aggregation of gatling logs, such as this one: https://github.com/codefiler/gatling-analyser
