Problems with StatD and Graphite with stats_counts tags

Question

I am having problems with stats_counts labels that whipser saves.

In my application, I run two kinds of indicators at the same time: Counter increment + Time. When I compare specific graphs of indicators: stats_counts + "count" of the timer, I get the same results for 4 hours, but when I look at a longer period - say, 12 hours, I see big differences.

I found this post: Tracking metrics using StatsD (via etsy) and Graphite, the graph graph doesn't seem to display all the data, but that didn't help me.

It seems to me that this is a configuration problem, but I can not understand what the problem is.

Here's what my configuration files look like:

storage-schemas.conf:

[stats] priority = 110 pattern = ^stats\..* retentions = 10s:6h,1m:7d,10m:1y [stats_counts] priority = 110 pattern = ^stats_counts\..* retentions = 10s:6h,1m:7d,10m:1y 

storage-aggregation.conf:

 [min] pattern = \.min$ xFilesFactor = 0.1 aggregationMethod = min [max] pattern = \.max$ xFilesFactor = 0.1 aggregationMethod = max [sum] pattern = \.count$ xFilesFactor = 0 aggregationMethod = sum [stats_counts] pattern = ^stats_counts\. xFilesFactor = 0 aggregationMethod = sum [default_average] pattern = .* xFilesFactor = 0.3 aggregationMethod = average 

As you can see, I have a stats_counts rule for aggregation, but it seems that it does not work so well, any idea that I am missing? perhaps the regex syntax I wrote there is not very good: ^ stats_counts.

Answer 1

The problem was that when I call a stop during the carbon caching process to refer to its memory with a newer configuration, it did not stop, so the configuration was old.

So, there is nothing wrong with the configuration described above. Just make sure your configuration is actually absorbed by the carbon process - with the whisper-info.py command:

whisper-info.py your_metric_file_name.wsp

Note that the problem with disabling the carbon cache looks: https://answers.launchpad.net/graphite/+question/135957