Closed peterbe closed 6 years ago
cc @bqbn
Even better, markus already has a backend for this! https://github.com/willkg/markus/blob/master/markus/backends/cloudwatch.py
That PR does everything we can do (hopefully!) to enabled Datadog integration in Lambda. We just need to remember to set LOG_METRICS=cloudwatch
in our Lambda environment.
@bqbn This is what you did to make this work in Stage Lambda.
At the moment, Lambda on Prod is so old that it doesn't even use the markus
metrics library.
I don't know what you need to do in cloudops-deployment
so that it's enabled for Prod too so that we can close this issue.
I'm closing this because, correct me if wrong, we now have the lambda metrics in Datadog for -stage, which means once the latest version (v1.3.2 as of this comment) is released to -prod, we should be able to see the same metrics in Datadog for -prod as well.
We can re-open this issue if, when the latest version is released, and we still don't have lambda metrics in Datadog for -prod.
According to the documentation it should be possible: https://docs.datadoghq.com/integrations/amazon_lambda/
When I go to https://app.datadoghq.com/account/settings#integrations/amazon_lambda I think what that page is trying to say is that it's enabled.
This issue is two things:
stdout
writes (using our current metrics infra) so that the relevant metrics can be sucked up by the Datadog/Cloudwatch/Lambda integration.