Throttle events

[whysthatso]

heya,
first of all, thank you for all your work in the open, it's been delightful to follow.

i've just set up a legacy app with honeybadger that is currently on airbrake in order to get a second opinion and be able to correlate logs also with errors.

first time i'm using insights, and after a day of running it on staging, i see the quota being eaten by solid_queue stats events. solid_queue is used with a puma plugin worker inside a container, so by default all goes into one big log stream currently. i think that is quite a common starter situation.

I can see there are being about 10 stats events collected every minute or so, and i'm wondering if there is a reasonable way to reduce this granularity?

[roelbondoc]

Hey there @whysthatso,

You can configure the data aggregation frequency by setting the following in your honeybadger.yml configuration file.

insights:
  registry_flush_interval: 120

The registry_flush_interval is the amount of time in seconds for the gem to collect and aggregate data before sending them to Honeybadger Insights. This would increase the interval of sending those stats to every two minutes instead.

Thanks for your feedback about this being a common starter situation. I'll look into seeing if we should increase the default interval.

Here is more information about that in our docs.

If you don't need that data, you can also disable the collection of those stats by setting the following:

solid_queue:
  insights:
    cluster_collection: false

Note that solid_queue is the top level key for this config and not insights as in the previous example. This is because this is a plugin specific setting.

Let me know if that helps your situation. Feel free to reach out with any questions or concerns. You can also contact us at support@honeybadger.io, and I can take a closer look!

Roel