Unrelated topic, I’m having a bit of trouble using S3 storage for a flow. Nothing seems to get created at the key I specify, but I’m not getting any errors.

Hi there, I am trying to build a "flow of flows", where two flows (a and b) are run by a parent flow (p). I have a cron clock which will run 5 times a day at specific times. I need to pass two varying parameters (t1 and t2) to flows a and b at the beginning of each schedule. These parameters represent timestamps where, the first time everything is scheduled, t1 will be

'0000-00-00 00:00:00'

and t2 will be

time.now()

. The next time schedule runs, t1 needs to be the previous value of t2, and t2 will once again be

time.now()

. I can see that I may need to use a task to compute the parameters within the script of the parent flow (similar to here) and use the

StartFlowRun

task (as seen here), but I'm not quite sure how to put it together. Any help would be appreciated, thank you!!

HI there, i have issue while running a workflow through ecs agent ,

An error occurred (ClientException) when calling the RegisterTaskDefinition operation: Container.image should not be null or empty

. this is run time error while trying to execute flow . any pointers

Hey community, I have previously worked with Dockerstorage for all my flows but as my repo has grown it has slown down my CI process, so I am switching to S3 stroage. I have a docker image baked in with all the dependancies and files etc for the repo. I have set up my storage like so:

S3(
            bucket="es-prefect-flows-staging",
            stored_as_script=True,
        )

Now after registering my flow I am getting an error after the flow is downloaded: in 🧵

Hello, I have 2 questions about Prefect Cloud: • Is there a list somewhere that detail the information sent to Prefect Cloud from Agents? • Is there documentation around Okta SSO implementation with Prefect Cloud? (more specifically when I when to have multi-tenancy) Thanks in advance for the help :)

Hey folks, Haven’t been able to figure this one out by reading the docs / threads here. I am trying to process data for sites in a date range one day at a time (with a task running for each combination of date and site_id). My current solution uses

map

to process the data for each site individually, but does so for the entire range at once. The complication arises from the fact that the date range is specified as a parameter, and the site_ids are the result of a task that runs.

I'm trying to update the flow schedule using graphql, there is an option in UI under setting where I can modify schedule. I want something similar in graphql

Hello everyone, I noticed something strange today and I don't know if it was a bug or not. I had registered a Flow using the API and started an agent at the same folder using the CLI. The Flow was schedule to Run on the Prefect Cloud and I did some "Quick Runs" to test/debug. Everything was running fine, however, the VM that I'm running the Agent was shut down for maintenance on the weekend, and I restarted the agent again (now in a different folder than before) an received the following error:

Failed to load and execute Flow's environment: FlowStorageError('An error occurred while unpickling the flow:\n  ModuleNotFoundError("No module named \'Submodules\'",)\nThis may be due to a missing Python module in your current environment. Please ensure you have all required flow dependencies installed.',)

The error was solved after shutting down the agent and starting again at the same folder of his first execution. I think this is related to the agent metadata, is this behavior normal?

When exactly do I need a

RunConfig

object? Do I need one if I’m already specifying an Executor?

Hey all, what is the best practice for creating variables (i.e. date stamp) that you want to re-use across tasks? I can't just declare the variable as the date stamp doesn't update after the initial registry.

Hey everyone, I'm getting a weird error when trying to run flows with the docker agent against the cloud backend. The error message asks me to make sure I'm authenticated. However, when I run

prefect auth list-tenants

I get my tenants, so if feel like I should be authenticated. I'm kind of stumped, maybe someone here has an idea?

Traceback (most recent call last):
  File "/usr/local/lib/python3.7/site-packages/prefect/client/client.py", line 738, in _request
    json_resp = response.json()
  File "/usr/local/lib/python3.7/site-packages/requests/models.py", line 900, in json
    return complexjson.loads(self.text, **kwargs)
  File "/usr/local/lib/python3.7/json/__init__.py", line 348, in loads
    return _default_decoder.decode(s)
  File "/usr/local/lib/python3.7/json/decoder.py", line 337, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/usr/local/lib/python3.7/json/decoder.py", line 355, in raw_decode
    raise JSONDecodeError("Expecting value", s, err.value) from None
json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/usr/local/bin/prefect", line 8, in <module>
    sys.exit(cli())
  File "/usr/local/lib/python3.7/site-packages/click/core.py", line 829, in __call__
    return self.main(*args, **kwargs)
  File "/usr/local/lib/python3.7/site-packages/click/core.py", line 782, in main
    rv = self.invoke(ctx)
  File "/usr/local/lib/python3.7/site-packages/click/core.py", line 1259, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/usr/local/lib/python3.7/site-packages/click/core.py", line 1259, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/usr/local/lib/python3.7/site-packages/click/core.py", line 1066, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/usr/local/lib/python3.7/site-packages/click/core.py", line 610, in invoke
    return callback(*args, **kwargs)
  File "/usr/local/lib/python3.7/site-packages/prefect/cli/execute.py", line 53, in flow_run
    result = client.graphql(query)
  File "/usr/local/lib/python3.7/site-packages/prefect/client/client.py", line 550, in graphql
    retry_on_api_error=retry_on_api_error,
  File "/usr/local/lib/python3.7/site-packages/prefect/client/client.py", line 454, in post
    retry_on_api_error=retry_on_api_error,
  File "/usr/local/lib/python3.7/site-packages/prefect/client/client.py", line 744, in _request
    ) from exc
prefect.exceptions.AuthorizationError: Malformed response received from Cloud - please ensure that you are authenticated. See `prefect auth login --help`.

Teams question If you have several teams, how do you register flows from the same server to a different team? I do not see a

team

option in

flow.register

. is it when you create the project in the cloud UI? what happens if one team create a project with the same name as a different team ?

Is there a way to assign a task result handler from an imported task? Like the equivalent of this:

from prefect import task
from prefect.engine.result.local_result import LocalResult
from my_prefect_tasks import imported_task

@task(result=LocalResult())
def assigned_result_task(x):
    return x

imported_task.result_handler = LocalResult()

when i run flows locally, my logging works fine, but if i run flows in prefect cloud, i get zero logging output. am i doing something blaringly wrong?

Hi all. I am seeing many of our jobs fail with this failure mode

requests.exceptions.RetryError: HTTPSConnectionPool(host='<http://api.prefect.io|api.prefect.io>', port=443): Max retries exceeded with url: /graphql (Caused by ResponseError('too many 429 error responses'))

Ours jobs are quite parallel, and wonder if there is anything that must be resolved, either user or server side that can enable this parallel execution.

Is there a way to extract the number of remaining tasks in a flow from something like the

prefect.client.client.FlowRunInfoResult

that's returned by

client.get_flow_run_info()

?

I’m getting

Unexpected error occured in FlowRunner: ModuleNotFoundError("No module named 'prefect'")

when I try to run a flow after making a small tweak to the pod spec. Anyone seen that before?

Hi, I have a job configured with ECS agent and flow runs in Fargate. It works fine, but when 10 jobs got triggered simultaneously (externally with lambda), couple of the jobs failed with

State message: An error occurred (ClientException) when calling the RegisterTaskDefinition operation: Too many concurrent attempts to create a new revision of the specified family.

Any advices?

I’m encountering strange behaviour when trying to invite Team Members to Cloud. I select the role as User, but the menu seems to randomly change to Administrator without me doing anything.

Getting some odd behaviour with a task, runs fine and finishes successfully - but is then started again? The task timer duration stops but is printing to logs. Using

DaskExecutor

💃 Just arrived!

Hey Prefect team, I'm having some difficulty understanding the proper arguments to modify/update an action. I am running: mutation { update_action (input:{ action_id: "<Action ID>" name: "CH4 Prefect Agents Down" }) { success } } But in return i get an internal server error; i just want to rename the action specified in the ID • nevermind resolved - just resupplied the config argument

Hmm, now I’m getting

ERROR - prefect.FlowRunner | Unexpected error: TypeError('code() takes at most 15 arguments (16 given)')

Is this potentially because I’m running Python 3.9?

We are upgrading from 13.7. Getting an error when run flows

Error downloading Flow from S3: No module named 'boto3'

Used the

prefecthq/server:core-0.15.3

image. Is there something for extras or something we are missing?

Has anyone had success with a single agent deployment handling multiple environments via labels? (more info in thread)

Hey all, I was wondering if there’s anyway to use Prefect (I have the services self hosted on a kubernetes cluster) with multiple tenants. I see the

tenant

field in the database, but could find a way to make the UI tenant specific, neither to register the flow for a specific tenant. Any ideas?

How do I setup a set of mapped runs to not fail when a single on fails?

Hi there, I have a Redis kv store set up and running, and I would like to set and get values for timestamps using the conditional ifelse task. I am running into some troubles when trying to get the False case:

import time
import prefect
from prefect import task, Flow
from prefect.tasks.redis import RedisSet, RedisGet
from prefect.tasks.control_flow import ifelse

redis_get = RedisGet(...)
redis_set = RedisSet(...)

with Flow("flow_p") as flow:
    t1_get = redis_get(redis_key='t1_key')

    # true case
    t1_init = redis_set(redis_key='t1_key', redis_val="0000-00-00 00:00:00")
    # false case
    t1_set = redis_set(redis_key='t1_key', redis_val=time.strftime('%Y-%m-%d %H:%M:%S', time.localtime()))
    
    ifelse(case(t1_get, None), t1_init, t1_set)

flow.run()

Running once then doing

RedisGet(...).run(redis_key='t1_key')

gives

b'0000-00-00 00:00:00'

but running again gives the same value. Any ideas why I am not getting the False case on the second run?

1. Is there any way to setup Flow Concurrency control to simply skip scheduled runs if another flow is already running with the same labels? 2. Is it also the case that the Agent has to be tagged with the Flow Concurrency control label as well? So presumably if I have 20 different Flow Concurrency labels, all 20 need to be attached to the Agent?

I am doing this python prefect_pipeline.py (this creates a flow and registers() and runs() it. when I do this , i only see that the prefect task is only registered in the prefect UI but actually not running in the agent but just in my local machine from where I invoke the python prefect_pipeline.py code..