Question about retries: if I implement the retry at flow level, is there a way to execute only the failed tasks on the next retry?

See more about results here. Note that flow runs and caching do require results persistence.

Share code for module 102:

room 5

room 7

Is there a way to have implicit task dependencies in the same way Prefect 1 did ? e.g. something like

@flow
def my_flow():
    a = task_A()
    b = task_B(upstream_tasks=[a]) # I really really need B to run after A but B doesn't need result from A

room1 - session 2 thread pool

Is it expected that cached tasks not show up in the flow graph UI? In the example below,

hello_task

does not show up (hits the cache)

Rest api documentation, it allows you to have the most control on what happens with the orchestrator

can an automation execute custom code? or should it be a Block only?

lab 103 instructions: • Use a previously made flow • Create a Cloud profile (or use an existing one) • Authenticate for the CLI • Make an email notification automation for flow run completion • Run the flow a few times from the CLI • See the event feed in the UI • Stretch: create a Slack notification automation

Main room

room 1 thread

room 9 thread

room 8 thread

Recording from today: https://us06web.zoom.us/rec/share/7wPEOmxMIL2wQVKR-rhrvIbC5g49sPSQdxUAsuurmXjCa6x_LYsgFxEmobe3Cbkf.mGgQH1XYsfRCZR4D Passcode: vc0ct$qr

Thank you for all your hard work today! sonic Following up on a few questions:

I filed an issue in the GitHub repo for no link to persisted flow run result in the UI.

Caching docs have details you might be curious about.

Logging: You can log output to a file from your terminal as you would with other programs. e.g.

python my_flow.py 2> test2.txt

If you want to add additional loggers, there is a prefect setting for that. You can add it to your Prefect profile or create an environment variable. Docs here. Example discourse post here.

Question about retries: if I implement the retry at flow level, is there a way to execute only the failed tasks on the next retry? Slack Conversation

Good day! 😎 We’ll get the Zoom room opened at 8:55am ET to start at 9am ET.

can share the slides ?

Day 2 slides:

@Javier Ochoa if i remember correctly, in prefect 1 you could associate a set of parameters with the schedule itself, and then that schedule with a flow in prefect 2 you could create a new deployment for each unique schedule you'd like to apply and do the same for parameters (overriding values as needed)

we have an ssl certificate issue when trying to run worker, room 8

We’ll get back together at 8 minutes till 10am ET. Feel free to take a quick stretch break.

Override default parameters on the deployment UI doesnt show in quick run, it still asks to provide parameters, is it a bug ?

room 6

Prefect worker (or agent) as background service - short answer, you can use systemd on linux.