New RunLLM response generated: >>>*# Question* Hi everyone! I have a workflow that has subworkflows - these workflows have maptasks. in some cases: some arraynodes get stuck on "running" status preventing next-in-queue tasks/arraynodes to start despite that the logs show that the process finished. both the master workflow and subworkflows have FAIL_AFTER_EXECUTABLE_NODES_COMPLETE failure policy maptasks are set with concurrency=2, min_success_ratio=0.1 Image flytekit version: 1.12.0 cluster flyte version: 1.13.1 any idea on why that happens ? # Answer This issue might be related to a known problem where tasks or array nodes appear stuck in a "running" state even though they have completed. This can occur due to eventing issues where the status is not correctly updated in the Flyte console, as noted in Flyte Issue #5335. Additionally, there have been reports of similar behavior when using map tasks with specific configurations, such as

min_success_ratio

, which can lead to unexpected task states (Flyte Issue #5201). Would you like more detailed information on potential solutions or workarounds for this issue? Sources: - Flyte Issue #5335 - Flyte Issue #5201 Slack Link https://.slack.com/archives/CP2HDHKE1/p1729588929.510089 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.