May 2019
Intermediate to advanced
442 pages
11h 36m
English
The most common pitfall when starting out with PromQL is not choosing the right function for the data type (such as counters, gauges, or histograms) or vector type. Even though this information is pointed out in the Prometheus documentation, there can still be some confusion as there are conceptually similar named functions and aggregators: rate and deriv, increase and delta, quantile and histogram_quantile, sum and sum_over_time, among others. Fortunately, in cases where there is a mismatch of vector types, the expression evaluation will fail and let you know what is wrong; for a mismatch in data types, such as providing a counter to a function that expects a gauge, the expression might evaluate ...
Read now
Unlock full access