Flushing the same metric multiple times in the same second overwrites previous values

If you increment a counter, flush the message, and then before the UTC timestamp has rolled into the next second, increment the same counter with the same tags (and flush again), only one of your two increments will show up in datadog. I think Datadog has some de-duplication / unique indexing that only tracks a unique metric per second. I think this library would need to keep in-memory up to 1 second of messages, to be able to track and send the right stats.

Start by tracing how counters are incremented and flushed through the DataDog HTTP API, focusing on the UTC-second timestamp and identical tags. Reproduce two same-second flushes for one counter, then define the fix by verifying that both increments are represented rather than overwritten by Datadog.

Written by the indexing model from the issue text.