Hey y'all, I've run into this issue where my task will be created & submitted and then... nothing. It's just stuck in Pending. I'm trying to bring Prefect to my company for data engineering, and I really need to figure this out to move forward. Thanks! 14:00:11.934 | INFO | Flow run 'dazzling-lorikeet' - Created task run 'read_file-7f2c7640-0' for task 'read_file' 14:00:11.935 | INFO | Flow run 'dazzling-lorikeet' - Submitted task run 'read_file-7f2c7640-0' for execution. I can read the file if I copy-paste the code into the flow, so I know it's not an issue there. UPDATE: With no flows running, somehow my concurrency-tag said that it was full (3/3 active tasks), so this task never made it into the queue.

hello all! 👋 new prefect user here. We’re still trying to learn the tips n tricks surrounding prefect, one question I had was how can we manually stop flow runs via the UI? I might be missing something, but I can’t seem to find it on the UI

is there any easy way to schedule a flow for say every 30m at the beginning of the day, and then when the flow succeeds once skip the rest of the runs for that day?

Hello community 👋, I have a personal project in mind that i want to orchestrate a 😛ython: python script to run once daily on Prefect 😛refect:. I want to stick to best practice as possible, and i read that you shouldn’t be executing processes on the same instances as your orchestrator. Does it make sense for prefect to call :gcp: GCP Cloud function to execute my python script? Im open to ideas!

Hi community, what would be a suggested/recommended setup to be used with azure as host and Prefect as keystone ; about 10 python scripts are executed daily 8 are lightweight and 2 of them do a lot of heavy lifting , they are changed with lots of frequency so maintenance overall cant be too complicated but for cost optimization some sort of elasticity would be most appreciated as the requirements of the heavy scripts are 10x the small scripts hardware reqs. thanks a bunch in advance

Hi prefectionists, first of all great work on a promising product! I am looking for a solution rec for this use case. Suppose there are events that starts at unknown times. Once it is detected that an event has started a flow is looped with some persistent state until it's detected that the event has ended. There can be multiple events happening at once, so there can be multiple loops but only as many as there are events. Therefore, only one loop for each event, as there are no safeguards for multiple loops working on the same event with the same state and inserting identical data. Normally, I would have a repeating event monitoring job that detects the start of events, publish the job with initial state into a work queue, and job consumers that carry out the job, and re-publish into the queue once completed with latest state so the next iteration can begin. The same job can detect end of event and complete without publishing back into the queue. What would be the ideal setup using prefect? thanks in advance for reading the long rant!

Hi community, I just want to undertand is there a way I can interact with prefect apis (not the cloud version) to handle multi-tenancy ? Can I create workspaces in the open source using the prefect apis ?

Is there a workaround to do

prefect cloud login

without verifying ssl certs? I'm getting an ssl, unable to find local issuer cert error. I can resolve it with IT it'll just take a week and a bit to go through the whole process. I'm just prototyping some stuff. I checked here: https://discourse.prefect.io/t/how-to-disable-the-ssl-verification-when-setting-up-a-pr[…]verify-failed-unable-to-get-local-issuer-certificate/597/2 But the solutions didn't seem to work. I'm using windows btw

Also if I do put in a ticket with IT what domain do I need to whitelist *.prefect.cloud ?

Hi Prefect Team, first a thank you to the wonderful job you have done creating this platform! I have a question about heterogeneous flow run: 1. Is there any way for a user to create a flow that runs on different type of hardware in the queue? I need to create a flow that partially run on CPU kubernetes queue and partially on GPU enabled nodes. There doesn't seem to be a way to do this and each flow can only run on pure CPU or all with GPU. 2. If not, what's the best practice for doing this? Should I break a flow into separate subflows and deploy them separately and run each step by hand? Or can I trigger this automatically by having a task run other deployed flow in a different queue. 3. Is there any plan to add per task tagging system for agents so when an agent pulls from a queue it only pulls tasks with a certain tag

Is this a supported pattern in prefect 2? • Task A generates some output • Use a remote storage for persistence • Task B retrieves output from storage I want to be able to: • Control the filename in azure blob store (right now prefect gives it some random nonsense name) • If the file already exists, skip task A

Hi - its been toted that prefect could be used to run arbitrary workflows in languages other than python (via docker or aome other means). Maybe it's a rare use case but I haven't seen an example of this. Are there any toy examples around?

I'm just upgrading to prefect 2.0 and I' finding the secrets don't work when I try to get() them. There's a post in the community forum, but it's not resolved. Has anyone run into this problem

AttributeError: 'coroutine' object has no attribute 'get'

and resolved it?

Hi all, I am curious if there is a method in prefect 2 to allow for manual reading of persisted results files. I see that PersistedResult object has a .get() method, however from what I gather this is meant to be used behind the scenes. If I can use it manually, could someone show an example?

Hey everyone! Just wanted to share something I found interesting. When submitting many

async

tasks, either via

.submit

or

.map

, I've found that using a

SequentialTaskRunner

causes the flow to run much faster than using

ConcurrentTaskRunner

. I'd like to know more about this behavior. Any feedback is appreciated!

Hello all, I have been looking for if prefect 2 is able to leverage different container for different tasks. At Datateer we are leveraging kubenetes as our deployment model. There was a post made on the discord asking the same question. I also found on the discord for prefect 1 that there are ideas that lead me to the kubernetes job task. This lead me to find this prefect-kubernetes package which provides support for prefect 2. I also know that there is the native ability to run sub-flows in a task runner through dask. My question is if there is an accepted standard for managing tasks that one would want to run in an isolated container to better manage dependency conflicts and design runtimes for the tasks that are optimized for the needed resources. The latter would help with reducing runtime costs.

Hi all, I think I'm doing something simple and mistaken. I'm using Python 3.9.6 on Windows 10, and Prefect version 2.8.4, I have the simplest code in a file called dev.py from the 1st tutorial example from the site. I'm using a Python venv too

from prefect import flow

@flow
def my_favorite_function():
    print("What is your favorite number?")
    return 42

print(my_favorite_function())

And when I run

python dev.py

it gives me an error

RuntimeError: Cannot create flow run. Failed to reach API at <http://127.0.0.1:4200/api/>

I tried changing the code to:

from prefect import flow

@flow
def my_favorite_function():
    print("What is your favorite number?")
    return 42

if __name__ == "__main__":
    print(my_favorite_function())

And it still gives me the same error. Am I missing something? I'm sure I've run this code before when trying to learn

[THEORETICAL QUESTION] Hi everyone, new to DBT / Prefect and Data Engineering. I am currently building my first DE capstone project where I am pulling data from an API, and pushing it to BQ. Orchestration happens with Prefect. I want to use DBT to partition and prepare the data in BQ for Looker. I am having issues wrapping my head around how I can make my code reproducible for others on GitHub as I am using DBT cloud. I found repo from Anna but it does not fully answer my question on how to make the code reproducible (especially in terms of DBT credentials). What would be the best practices here? Would I need to explain them how to setup the DBT cloud as well or is there a better way?

[Beginners questions] [VM coordination with Prefect Cloud] [Work pools] [Agents] Hey there, I successfully deployed a prefect flow and it perfectly pulls data from an API and uploads it to GCS and then to BigQuery. But when it comes to scheduling I am missing understanding. I use a VM from Google and host my code on GitHub. When I am running the flow from my VMs CLI: 1. I start the default work pool via

prefect agent start -q 'default'

2. I call my flow

python my_flow_file.py

I would like my flow now to run independently without me needing to do (1) and (2). I scheduled my flow to run every Saturday, but when I checked, today, it was in status

late

and was waiting for a agent to pick up the run. How would I start an agent when I am not in VS code, starting it via the CLI? I assumed it would be started via the Prefect Cloud or my VM that was running.

[DBT] [Prefect] Thanks for the relentless support everyone. I am amazed by Prefect and really want to dive deeper into it! I am trying to orchestrate my dbt job with prefect. Following these instructions I was able to set up the credentials block. Unfortunately the job is not executed and it throws no error 🧵 - somebody with a similar setup that encountered this issue as well?

Hi all, I need help setting up deployments using a docker container as infrastructure (rather than the local environment in which the agent is running). For context, the deployment works when I don't specify infrastructure. The error I receive is: 14:31:42.716 | INFO | Flow run 'enigmatic-octopus' - Downloading flow code from storage at 'generic_wholesale_cert_folder' 14:31:43.776 | ERROR | Flow run 'enigmatic-octopus' - Flow could not be retrieved from deployment. Traceback (most recent call last): File "<frozen importlib._bootstrap_external>", line 839, in exec_module File "<frozen importlib._bootstrap_external>", line 975, in get_code File "<frozen importlib._bootstrap_external>", line 1032, in get_data FileNotFoundError: [Errno 2] No such file or directory: 'sql_queries/sql_query_pull.py' The above exception was the direct cause of the following exception:

Hello, I need help with the following stack trace:

Flow could not be retrieved from deployment.
Traceback (most recent call last):
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/prefect/engine.py", line 247, in retrieve_flow_then_begin_flow_run
    flow = await load_flow_from_flow_run(flow_run, client=client)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/prefect/client/utilities.py", line 47, in with_injected_client
    return await fn(*args, **kwargs)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/prefect/deployments.py", line 159, in load_flow_from_flow_run
    await storage_block.get_directory(from_path=deployment.path, local_path=".")
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/prefect/filesystems.py", line 553, in get_directory
    return await self.filesystem.get_directory(
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/prefect/filesystems.py", line 310, in get_directory
    return self.filesystem.get(from_path, local_path, recursive=True)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/fsspec/asyn.py", line 113, in wrapper
    return sync(self.loop, func, *args, **kwargs)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/fsspec/asyn.py", line 98, in sync
    raise return_result
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/fsspec/asyn.py", line 53, in _runner
    result[0] = await coro
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/fsspec/asyn.py", line 561, in _get
    return await _run_coros_in_chunks(
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/fsspec/asyn.py", line 269, in _run_coros_in_chunks
    await asyncio.gather(*chunk, return_exceptions=return_exceptions),
  File "/usr/lib/python3.10/asyncio/tasks.py", line 408, in wait_for
    return await fut
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/gcsfs/core.py", line 1266, in _get_file
    await self._get_file_request(u2, lpath, callback=callback, **kwargs)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/decorator.py", line 221, in fun
    return await caller(func, *(extras + args), **kw)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/gcsfs/retry.py", line 115, in retry_request
    return await func(*args, **kwargs)
  File "/home/datasc/tags-scripts/venv/lib/python3.10/site-packages/gcsfs/core.py", line 1253, in _get_file_request
    f2.write(data)
OSError: [Errno 28] No space left on device

Its deployed and running on GCP

I think your last line there tells you what you need to know

I am trying to run a deployment via Docker. However, I receive the following error:

FileNotFoundError: [Errno 2] No such file or directory: '/opt/prefect/flows'

Which is surprising because I don't use the above location to reference the flow. This is what my Docker file look like

FROM python:3.8-slim-buster

ARG PREFECT_API_KEY
ENV PREFECT_API_KEY=$PREFECT_API_KEY

ARG PREFECT_API_URL
ENV PREFECT_API_URL=$PREFECT_API_URL

ARG GCP_DATASET_NAME
ENV GCP_DATASET_NAME=$GCP_DATASET_NAME

ARG GCP_DATASET_TABLE_NAME
ENV GCP_DATASET_TABLE_NAME=$GCP_DATASET_TABLE_NAME

ARG GCP_PROJECT_ID
ENV GCP_PROJECT_ID=$GCP_PROJECT_ID

ARG GCP_REGION
ENV GCP_REGION=$GCP_REGION

COPY poetry.lock .
COPY pyproject.toml .

RUN pip install poetry --trusted-host <http://pypi.python.org|pypi.python.org> --no-cache-dir
RUN poetry config virtualenvs.create false
RUN poetry install --no-root --without dev


RUN mkdir scripts
copy scripts/ scripts

RUN mkdir config
COPY config/ config

RUN mkdir -p dbt/xetra
COPY dbt/xetra dbt/xetra

Any idea why Prefect is looking for the flow in the /opt/prefect/flows directory? I am running this via the Cloud

I was at PyCon last month and met some folks from Temporal (https://temporal.io/) who work on workflows similar to Prefect. I looked at their documentation but could use help deciphering their advantage over Prefect. Does anyone know the pros/cons of Temporal vs. Prefect? I ask because we're using Prefect to automate our data pipelines and even some of our application data refreshes (replacing celery workers). I'm wondering what advantage Temporal would provide over Prefect in this regard.

Hi, Is it possible to remove duplicate todos with prefect ? for now I'm using Queue.queue and thread locks to have 1 jobs of type with url = https://google.fr and 1 job with this parameter in queue max ?

Hi I have an issue when writing to a google cloud storage, when I use the

upload_from_path

on my code you will see that I am passing the same variable path as the

from_path

and the

to_path

but for some reason prefect changes the structure of the

to_path

variable, here is the code I have that builds the path:

@task()
def write_local(df: pd.DataFrame, color: str, dataset_file: str) -> Path:
    """Write DataFrame out locally as parquet file"""
    Path(f"data/{color}").mkdir(parents=True, exist_ok=True)
    path = Path(f"data/{color}/{dataset_file}.parquet")
    df.to_parquet(path, compression="gzip")
    return path


@task
def write_gcs(path: Path) -> None:
    """Upload local parquet file to GCS"""
    gcs_block = GcsBucket.load("zoom-gcs")
    gcs_block.upload_from_path(from_path=path, to_path=path)
    return

you can see in the second task

write_gcs

both of the paths are the same variable called

path

and that is just a path structure that has originally this value:

'data/yellow/yellow_tripdata_2021-01.parquet'

. The prefect flows runs, but after it runs, in the details of the flow we can see on the second picture I am attaching it changed the text structure of the path for GCS to:

'data\\yellow\\yellow_tripdata_2021-01.parquet'

, no idea why this is happening and because of this you can see in the picture 1 that it saves the file with that weird name instead of creating the folders in GCS, any help on maybe why this is happening?

Hi All, I use the persist_result for my tasks, however I have two agents and when retrying flow it may happen that the flow no longer runs on the same agent. So I can't use local storage (PREFECT_HOME/storage) I don't want to use s3, is it possible to use prefect's blocks to store these results? or even better, use the postgres DB? Thanks

Hi everyone, I am running into some issues this morning with some of my deployments. I think it may be related to low disk space on our server but our deployments/ flows are failing with no logging. There is a State message to the right that says "Submission failed. FileNotFoundError: [Errno 2] No such file or directory: '/tmp/tmp9zwn92g4prefect". I think this has something to do with the space issue that we are working on addressing but I am curious why there would be no logging?

Has anyone set up a Zap in Zapier to trigger a prefect deployment via webhook? I'm having trouble figuring out how to authenticate a post request from Zapier securely