feat: Do not trigger the race flush if an ongoing continuous flush is… by astuyve · Pull Request #828

feat: Do not trigger the race flush if an ongoing continuous flush is… by astuyve · Pull Request #828 · DataDog/datadog-lambda-extension

Conversation

In rare cases a function will begin a continuous flush then become idle for a while. If and when it wakes back up, the race flush timer may trigger causing a pretty painful resource tie-up as the race flush will begin along with a new invocation as well as potentially another flush redrive/continuous flush. I noticed this can cause occasional timeouts.

		//Periodic flush scenario, flush at top of invocation
		if current_flush_decision == FlushDecision::Continuous {
		if current_flush_decision == FlushDecision::Continuous
		&& !pending_flush_handles.has_pending_handles()

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This prevents multiple flush tasks from piling up

		}));
		}

		while let Some(result) = join_set.join_next().await {

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Previously this code would await each request serially. At the time I didn't expect that to be an issue because it was being called from another task but ultimately it does slow things down as the final flush task was driving these serially

astuyve marked this pull request as ready for review

September 16, 2025 16:17

Conversation

Choose a reason for hiding this comment

Choose a reason for hiding this comment

2 participants