trimmed count in Cloudwatch dashboards for http status codes graphs shows unexpected results

0

We collect http status data via cloudwatch metrics and try to build dashboards widgets showing this data as graphs.

If we compare this metrics sample_count with the parallel collected log lines via Logs Insights stats count both sums are equal.

To form separate graphs for each https status codes or codes group (e.g. 200, 5xx, etc) we thought to use trimmed count. We assumed that TC(199:299) should show the number of all data points with a value from 200 till 299. We also expected to get always a integer back from this function. Instead we see floats as result.

We tried sample code and generated 2 status codes 500 within a set of roundabout 3460 other status codes and got from TC(499:500) a 0.42 back.

Is there a way to control what really is stored within a unique metric record? How does the trimmed count function really work?

No Answers

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions