Hello, I need to rename my sub_flow on runtime. I figured out how to do this with local Orion. I used a small hack from https://gist.github.com/crscheid/10dbf291779667831b856fab9cc9745b But the problem is following: I cannot rename flow with UPDATE method when I use cloud Orion. I have 422:

'{"exception_message":"Invalid request received.","exception_detail":[{"loc":["body","name"],"msg":"extra fields not permitted","type":"value_error.extra"}],"request_body":{"name":"d3f199ee-8bd5-4fbf-9424-a9a7f39c9d3b"}}'

Request in thread

Hi all. I am currently having a number of prefect deployments and work queues (prefect 2) that I would like to test out, and start the work-queues for their respective deployments. I know there is the CLI option (prefect work-queue start …) but is there a pythonic programmatic way to generate the list of work queues and start them through a function?

Hello! I am using a dask task runner for my flow and I was wondering if it is possible to use the dask cluster to run computations other than running prefect tasks. Is there a way to create a dask client and manually submit work to the cluster? thank you!

Hello All, I am currently trying to create a deployment that takes a parameter as a dictionary and not just a string. Is this possible within Prefect or is the only data you're able to provide as a parameter to a deployment a string? Best, Wright

I’ve got a question related to Snowflake and creating a credentials block. I’m attempting to use a private key, and can’t seem to get anything to work. Question: What is the right format for specifying the private key? I’ve tried a multiline string in the format of:

------BEGIN PRIVATE KEY-----

-----END PRIVATE KEY-----

I’ve tried taking the content between the blocks and making a long one line string and removing the

\n

characters. Each time I get the same InvalidPemFormat error.

Hi, I'm using Prefect cloud and I have created an API key. On the cli I do a

prefect cloud login -k

providing my key and selecting the proper workspace. Prefect tells me login has succeeded. If I then perform a

prefect block ls

I get a Client error '403 Forbidden'. I Prefect I have the Collaborator role. What could be the cause of this 403?

Reposting this from the prefect-ui channel - please consider participating. It should only take 5 minutes!

hey all, in prefect 1, i'm expecting a

None

value as a task result with

get_task_run_result

. i'm getting this error:

Task 'get_task_run_result[17]': Exception encountered during task execution!
Traceback (most recent call last):
  File "/usr/local/lib/python3.9/site-packages/prefect/engine/task_runner.py", line 880, in get_task_run_state
    value = prefect.utilities.executors.run_task_with_timeout(
  File "/usr/local/lib/python3.9/site-packages/prefect/utilities/executors.py", line 468, in run_task_with_timeout
    return task.run(*args, **kwargs)  # type: ignore
  File "/usr/local/lib/python3.9/site-packages/prefect/tasks/prefect/flow_run.py", line 239, in get_task_run_result
    return task_run.get_result()
  File "/usr/local/lib/python3.9/site-packages/prefect/backend/task_run.py", line 75, in get_result
    self._assert_result_type_is_okay()
  File "/usr/local/lib/python3.9/site-packages/prefect/backend/task_run.py", line 162, in _assert_result_type_is_okay
    raise ValueError(
ValueError: The task result has no `location` so the result cannot be loaded. This often means that your task result has not been configured or has been configured incorrectly.

is there a way to for me to access a task's result if it's

None

? My code works when i pass an

int

just seen there’s automation in Prefect Cloud, wanted to ask will it also be available for the open source version?

Hey guys, I'm a long time airflow user and considering prefect. How should I think about prefect? what are the main use cases I should have in mind when choosing Prefect over airflow?

Hi! we have installed prefect with graphql as agent in linux VM. The VM was restarted and the prefect is down now how do i restart prefect Please help me resolving this issue..

Hello, It seems like the signin link takes too much time to arrive in the email. Is it normal ?

[BUG?] [HELP NEEDED] Hi everyone, i'm having an issue with prefect on windows. On Windows, • We've bumped into an issue that prefect-gcp is actually converting the

/

(forward slashes) into

\\

double backward slashes during the upload tasks, • Which translates into the blob being named, for example, as:

foo\\bar\\foobar.txt

• Instead of actually:

foo/bar/foobar.txt

Which would be represented as "folders" in the GCS Web interface. We're not sure if the same issue happens on macOS or Linux We've tried not only the

upload_from_path

but also the

upload_from_file_object

https://prefecthq.github.io/prefect-gcp/cloud_storage/ Thanks in advance for any feedback

Hi, Can someone explain me how from we read configuration in prefect 2 from config.toml file earlier we were using

from prefect import config

what do we use in prefect 2 ?

Hi, I'd like to create automations (notifications) programmatically. I see the API documentation but I don't see any reference in the Python client (2.7.9). Is this still to be added?

Another automations question... I've configured an automation to send a slack notification when a work queue becomes unhealthy. If I then stop the agent process, the work queue does become unhealthy, but the notification is never sent/received. I know the Slack block works because I use it successfully to send a message when a flow changes state.

Hello folks, a possibly stupid question, but can't seem to figure this out by reading docs: I see that there's the new

PREFECT_API_ENABLE_HTTP2

to disable HTTP/2, but it's unclear to me if I have to pass that to agents, task runners (I use GKE), or both. I'm gonna set it in both but would want some guidance as to keep configs less polluted. 🙂 Thanks!

Cross posting from #prefect-getting-started in case someone has some suggestions

Hey Everyone, I have two pipelines/deployments A and B and they need to interact with each other, I have two scenarios that I can't seem to implement 1. Can two flows interact with each other and wait for each other to run a third flow, given that both flows are running separately and have separate deployments/pipelines? i.e. flow B has a dependency on both flows A and X. However, flow X is part of deployment B and is triggered by a different upstream process. Moreover, X and A don't necessarily run at the same time (~15 min gap) 2. Flow C is triggered by flow A on the 1st of every month, in normal circumstances, it would get executed, however, in occasional events, if it's Monday and also the 1st of the month, flow C should wait for B to finish, how does this work? I can just add an if-else or a similar implementation but wouldn't it mess up the radar (the visual flow representation) since the pythonic code can mess it up? Would really appreciate any lead! 🙂

How many times do you write "pythong" in a day?

Prefect v1 question - i sometimes see that a flow run or two will be stuck in a scheduled state - there are agents able to pick it up but it never goes from scheduled to submitted. Cancelling and restarting the flow run fixes it. A few times the issue is that they are stuck in a submitted state, but never go to running. Same things fixes those. Is there anything in particular that would cause this that I can attempt to change? Otherwise, are there any recommendations for how to write a graphql query to catch these and programatically cancel/restart them?

Hi! I want to use prefect2 to execute a flow and then wait for an irregular amount of time before executing it again. Is there a way to implement this?

I'm using result_storage to store the result of a flow ran with run_deployment in a S3 bucket. However, when I call .state.result(), I get a FileNotFoundError. Does anyone have an idea why this could be happening?

Wondering if I can run multiple instances of Orion server behind a load balancer for high availability?

Hi, is someone using python object to create deployments? I keep running into coroutine has no attribute apply. What am I doing wrong?

Hi, I am trying to save an ECS block from circleci deployment but getting a

401 Unauthorized error

. Works fine locally. What am I doing wrong? (details in next message)

Hello! I am using a dask task runner and I have noticed that if I create a bunch of tasks at once, some of them end up crashing and the flow fails. I was wondering if there is a way to initially create as many tasks as dask workers and only submit new tasks as previous tasks finish. Does prefect provide a method similar to

<http://dask.distributed.as|dask.distributed.as>_completed

https://github.com/dask/distributed/blob/main/distributed/client.py#L4955

Hey all, I'm exploring migrating from v1 to v2, and was looking to set up Notifications to a Slack channel. We have prefect-orion deployed via the helm chart, and currently have the Webhook url stored in an AWS Secret and would like to keep accessing it from there, but the only documentation I could find is for setting it up through the UI here. Is it possible to establish notification connections from outside the UI at all?

Hi Everyone, Is there an easy way to call prefect's rest API to kick off a deployment? There is an option to do this in python with the prefect library, but I want to run a deployment with javascript and an API call

hi, does anyone know if

prefect.context.flow_run_id

is accessible from a state handler, when the flow is cancelled and the state handler is invoked?