New RunLLM response generated: >>>*# Question* Hi everyone 👋 just getting started with Flyte and I have a question about task latency. I have a basic workflow consisting of one

ContainerTask

and one regular python task. Running it with

pyflyte run

locally finishes in under a second. But running with the flyte sandbox as a remote

pyflyte run --remote

it takes over 2 minutes according to the UI. Is this expected? It does eventually finish, Is there a timeout or something that can be configured? Details -> https://gist.github.com/perrygeo/05b438430ab776786bdc85185360c327 # Answer The latency you're experiencing is likely due to the overhead associated with "fast registration" when using

pyflyte run --remote

. This process involves packaging the code, uploading it to a blobstore, and then downloading and decompressing it at runtime, which can be time-consuming. To reduce this latency, consider using non-fast registration by pre-building a Docker image that contains your code. This approach eliminates the need to download and decompress the code at runtime, significantly speeding up task execution. Additionally, increasing the resource requests for your tasks can help mitigate delays caused by resource constraints. Would you like more details on how to implement non-fast registration or adjust resource requests? Sources: - Flyte task latency discussion - Flyte fast registration overhead - Flyte resource requests Slack Link https://.slack.com/archives/CP2HDHKE1/p1728516288.493979 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.