Why does DataDogProvider add_metric forceset a timestamp

[manzkel]

Hey,

we have been trying to send custom metrics with lambda powertools to datadog for a few days. While enhanced metrics were sent and received on Datadog, custom metrics did not show up.

After some debugging we found something peculiar in the DatadogProvider and would like to know if there is a reason behind the code and whether there are certain settings to be set to make it work.

Concretely, here

powertools-lambda-python/aws_lambda_powertools/metrics/provider/datadog/datadog.py

Line 117 in 85666bb

if not timestamp:

the setting of a timestamp is forced. Which in turn on datadog end forces the execution of this condition https://github.com/DataDog/datadog-lambda-python/blob/6e2e5d54b870bbe5d2b028e726fef1b5d8f70552/datadog_lambda/metric.py#L78

The extension does not support timestamps for distributions so we create a
a thread stats writer to submit metrics with timestamps to the API

Running this on our end did not succeed in sending / receiving custom metrics. Creating a custom Provider which does not override the timestamp however, allowed it to succeed.

Are there some settings we are missing to be able to use the Thread Stats Writer or is this an unwanted behavior?

Thank you!

[leandrodamascena]

Checking it now.

[leandrodamascena]

Just an update: I don't have the final answer here yet, but I hope to have it early next week! Thanks for your patience @manzkel

[manzkel]

Any news on this? 🙂

[leandrodamascena]

Hey @manzkel! I have some new findings and probably will submit a PR today. Sorry for the delay.

[leandrodamascena]

Hi @manzkel, I already have a branch working locally. We are releasing a version tomorrow, but I won't have time to include this change, so I hope to release it in the next few weeks.

BTW, can you confirm that change regarding DD_FLUSH_LOG was resolved?