Hi everyone, is there any suggestions to run task2 until task1 finished when they have no dependency?

with Flow('feature-factory') as flow:
    a, b = pre_task()
    task1.map(a)
    task2.map(b)

Would anyone be able to share what

ECSRun

is doing under the hood to run tasks? I have set a subnet for the service I am running my Prefect tasks in and have disabled public IPs, but when I call

ECSRun

my tasks are running with a public IP address in a different subnet. From reading the documentation for

ecs.run_task

it seems as though you are required to specify a subnet ID if you are running a task on Fargate, which I am. Therefore, I’m wondering what subnet ID Prefect is using and whether it can forced to use the default subnet ID of the service. I am aware that you can override the arguments passed to

run_task

by

ECSRun

, however I don’t want to do this as I want to make my config agnostic to the account that it is being deployed in (I am deploying the same infra to different AWS accounts and don’t want to hard code the subnet ID which is passed to

ECSRun

)

Hi! can I force prefect to re-run the entire flow when task failed? instead of continuing from the failing task on the retry?

Hey, quick question: Is there a way to access results as documented here https://docs.prefect.io/core/idioms/task-results.html when starting a flow with StartFlowRun?

Hi! I've been running a prefect server instance (version 0.14.11) in production for a few months, running a dozen flows, some of them every minute. The UI is getting extremely slow when clicking on a flow run or task run. Is this known to happen ? Should I delete old flow and task runs to limit the size of flow_runs and task_runs tables ? If yes, what's the best way to do it ? Thanks in advance !

Hello! I have Linux server that runs the agent that schedules new prefect flows' runs. And I have a GitLab CD pipeline that connects to this server on push to master and pulls master on that server. On push to master I have the latest version of the code on the server but I want the agent to start using the latest code while it keeps scheduling jobs with old code. Is there any kinda autoreload feature for prefect agent? Maybe some of you faced with that problem so I'm curious about how you solved it?

Hi folks, I have a Flow defined with a few tasks involved. Plan is to call a task from another task that’s called from a flow

With

context. Has anyone been able to successfully do that?

I run a prefect pipeline like this.

prefect run -n hello-flow --param args={'a':'alpha','b':bravo'}

Why cant i pass a dictionary to the param object instead of individual params

Error: No such command 'args ..blah

if I return a lot of data serialized with GCS results, is it expected that the only thing dask will transfer is the result metadata as opposed to the data itself? say for example I do

lots_of_dataframes_in_gcs = task(
    process_data,    
    result=GCSResult(
        PandasSerializer(file_type="parquet")
        )
    )
).map(input_files)
task(notify_data_is_ready).set_dependencies(upstream_tasks=[lots_of_dataframes])

I don't actually want the data passed to the downstream task, just to enforce the ordering. but based on my exploding daskworker memory it seems that the downstream task is actually collecting all of the outputs themselves(!), even though each transferred object shows up in the dask dashboard as a 48-byte

Cached

state object (so the memory is 99.999% untracked by dask) I can imagine a workaround that involves just skipping Results altogether and returning True or a file path explicitly, but it seems like this might be a bug and the

state.result

should be getting purged before we ship it around? @Chris White @Zanie or anyone else have thoughts on whether this is expected?

Folks, I need help with this warning message:

Flow run RUNNING: terminal tasks are incomplete.

Waiting for next available Task run at 2021-08-25T18:52:50.186382+00:0

that happens apparently in the end of the flow, but then it restarts the flow and execute everything again...

A couple of issues came up when I tried tor register a new flow: I develop my flows in a remote docker container, with local path /workspaces/prefect-dev/flows. When I try to register with working directory at prefect-dev using:

prefect register --project "some-project" --path flows/some_flow.py

my flow will get registered using the entire path(/workspaces....) . My agent searches for flows at ~/.prefect/flows, so I get a "Failed to load and execute Flow's environment" error. At the same time, even though I did not specify any labels, and there's no reference to any label whatsoever in any part of the source code, that flow gets created with one label that looks like an id of some sort (ie. 447c6079d9b5). My local agent is created using:

prefect agent local start --name $$(hostname) --no-hostname-label

Both of these issues came up with my latest flow, but I have used the same setup for a few months now without any of these issues. Any idea on what might be going on?

Hi everyone, I’m trying to re-write a flow written in the functional api into imperative api but I’m having a few issues with it: Functional api flow:

a_task = ATaskClass()

with Flow("Foo") as flow:
    name = Parameter('name')

    a_task = a_task(name=name)

Imperative api flow (so far):

a_task = ATaskClass()
name = Parameter('name')
flow = Flow("Foo")

flow.set_dependencies(
    task=a_task()
)

a_task.bind(name=name)

The error complains about the line where I’m trying to bind the flow param to the task:

raise ValueError(
ValueError: Could not infer an active Flow context while creating edge to <Task: ATaskClass>. This often means you called a task outside a `with Flow(...)` block. If you're trying to run this task outside of a Flow context, you need to call `ATaskClass(...).run(...)`

What’s the right way to go about this? 🤔 Some background info: our flows are getting quite large so we’re looking for ways to make them easier work with and more reusable. We considered using the flow-of-flows strategy but we don’t need to run any single part as an independent flow so splitting things up into X flows just means slowing down our development cycle since we need to register/re-register all those additional flows, instead of registering/re-registering the one big one. The imperative api seemed like a better choice since the fact that it’s so explicit makes it actually easier to read, specially when the flow is large, and we keep the registration to just that one flow. Many thanks!

Hey everyone, I'm currently learning my way around Prefect (and a lot of the other tools involved with Prefect) and tried to set up a simple hello task using GitHub storage. The task failed and resulted in this error in the log, could someone help me decipher it?

Failed to load and execute Flow's environment: GithubException(404, {'data': 'Not Found'}, {'server': '<http://GitHub.com|GitHub.com>', 'date': 'Thu, 26 Aug 2021 01:20:16 GMT', 'content-type': 'text/plain; charset=utf-8', 'transfer-encoding': 'chunked', 'vary': 'X-PJAX, X-PJAX-Container, Accept-Encoding, Accept, X-Requested-With', 'permissions-policy': 'interest-cohort=()', 'cache-control': 'no-cache', 'set-cookie': 'logged_in=no; domain=.<http://github.com|github.com>; path=/; expires=Fri, 26 Aug 2022 01:20:16 GMT; secure; HttpOnly; SameSite=Lax', 'strict-transport-security': 'max-age=31536000; includeSubdomains; preload', 'x-frame-options': 'deny', 'x-content-type-options': 'nosniff', 'x-xss-protection': '0', 'referrer-policy': 'origin-when-cross-origin, strict-origin-when-cross-origin', 'expect-ct': 'max-age=2592000, report-uri="<https://api.github.com/_private/browser/errors>"', 'content-security-policy': "default-src 'none'; base-uri 'self'; connect-src 'self'; form-action 'self'; img-src 'self' data:; script-src 'self'; style-src 'unsafe-inline'", 'content-encoding': 'gzip', 'x-github-request-id': 'EA90:30AB:16BB88:2112BC:6126EC50'})

good morning! What is the best way to parameterize Result location using Flow Parameter? Currently, I have, both at the flow and task level:

tsv_result_partial = partial(LocalResult, dir="./test_prefect", serializer = PandasSerializer("csv",
                                                                         serialize_kwargs={"sep":"\t", "index": False},
                                                                         deserialize_kwargs={"sep": "\t"}))

parquet_result_partial = partial(LocalResult, dir="./test_prefect", serializer = PandasSerializer("parquet"))

@task_no_checkpoint(target=search_result_df_file_name, result = tsv_result_partial())
def example_tsv_result_task:

@task_no_checkpoint(target=another_file_name, result = parquet_result_partial())
def example_parquet_result_task:

with Flow("test", result=LocalResult("./test_prefect")) as flow:
    param1 = Parameter()
    param2 = Parameter()

I want the outermost directory to be

result_folder/param1/param2

. How to achieve this?

Hi everyone — I’m have a flow that I want to run just one, immediately after I deploy it — is there a way to specify that on a schedule? I’m thinking about a

Schedule(clocks=[DatesClock([ pendulum.now() ])])

but I think the lag between running this and registering/deploying the flow would make it so that it doesn’t run, and I’m hesitant about hardcoding an “add 10 minutes” or something to it as doesn’t feel very robust

Hi ! I'm using Prefect Cloud and all of my previous runs have disappeared. They don't appear in the main dashboard, in the flows tab, or in the runs section of each flow. One of the flow that just ran is correctly shown though.

is there anyway to tie a readme to a flow while registering, or rn can you only do it via the UI?

It's possible to create custom states? Also, it's possible to aggregate (or search) by an error string (or by state) instead of debugging the log of each run on the Cloud?

Hey all, I feel like im missing something really simple

@task(result=LocalResult(location="initial_data.prefect"))
def root_task():
    return [1, 2, 3]


@task(result=LocalResult(location="{date:%A}/{task_name}.prefect"))
def downstream_task(x):
    return [i * 10 for i in x]


def rating_model_backtest_workflow(opts: dict):
    with Flow("local-results") as flow:
        downstream_task(root_task)

    flow.run()

I’m running this from the docs, I am expecting files to show up at

./initial_data.prefect

and the other location, but nothing appears, am I missing something obvious?

hey guys. trying to set the local configuration path, but is not being found. I’m doing as docs `export PREFECT__USER_CONFIG_PATH="path/to/config.toml"`I tried using both an absolute and a relative path and nothing. also tried settings the config (Slack webhook url) directly into the code, like this discussion and nothing, still getting the same

ValueError

where it can’t find Slack webhook URL

hi guys, thanks for the awesome tool! missing something really simple 😅 is there a way to access run parameters in the target template? i.e., how do i access the

company_name

run parameter when setting the target below (note, this doesn't work because of

dict object has no attribute company_name

):

@task(target="{parameters.company_name}/{today}-{task_name}", ...)

i tried

parameters.get("company_name")

instead but still no luck. must be a really simple thing i'm missing (sorry for the spam)

Prefect team Thank you for this great product …

Hey guys, I am exploring prefect in order to implement the various parts of an ML application (with multiple models) as a flow. Here are my queries regarding prefect tasks : 1. Are prefect flows a wise choice for productionalizing online streaming style ML models ? If yes what's the recommendation on mirroring the business logic as flows - is a flow of mapped flows standard or is creating a master flow to orchestrate other flow runs wiser choice? 2. Implementing an online streaming ML model as a task - right now I've been able to load the model inside of a task for prediction, but the model loading takes most of the time of the task run. Does prefect have mechanisms for me to persist the ML model in memory?

Hi all, has anyone else noticed that the prefect documentation website seems to become unresponsive after it's been open for several hours (or days?). What happens to me is that I can scroll and click links on the doc site, and the URL in the nav bar will change, but the page content will stay on the same content it was on previously when I left it and I'm unable to navigate away. To restore navigation features I need to refresh the page. Actually, I've also noticed the prefect UI interface also freeze up on me when it's left open for a long time as well (both cloud and self-hosted versions). More of a hard-lock there though where I can't seem to do anything. Curious if anyone else is experiencing one or both of these issues?

Hi there folks, how you doing? Github Actions deployment question I'm trying to improve our data team workflow with Prefect and our flows are using

Github Storage

. I'm facing a problem when "developing" the flows. Since we're using Github Actions, flows are registered to a

development

project when a Pull Request is opened (and updated). But

Github Storage

will use

main/master

as ref argument. This works fine for production, but won't do while I'm on a development branch. The workaround I've found is to use a pattern like this: • Github Actions Workflow with different steps for Push/Pull Request: ◦ Push to main: generate project folder and register the flow ▪︎ I use

prefect register --project $project --path $path

to do this ◦ Pull Request: register the flow with disabled schedule in

development

project ▪︎ I use the

if __name__ = "__main__"

pattern in order to pass a dynamic

GITHUB_HEAD_REF

to Github Storage ref argument. It's working, but I'm thinking if there is way just use the

prefect cli

for this and avoid writing boilerplate code. The thread has some code snippets and more details!

Hi folks. I’m trying to upgrade the Helm chart, to get it to re-create the default tenant for me. It tells me

'postgresqlPassword' must not be empty

even though I am passing it with

--set

on the

helm upgrade

command-line.

Is there any plan to create a task-library task for running Spark on kubernetes (as opposed to Databricks)?

Anyone ever seen

tornado.util.TimeoutError: Operation timed out after None seconds

before? I’m getting that in my Dask KubeCluster.

Is there a way to configure slack alerts from the flow definition (eg: like

flow.schedule

)? I've set them up in the UI a few times, but occasionally renamed flows or changed projects and lose failure alerts. 🙈

Hi guys. tell me, on my computer .py flow.run is executed without errors. But if I register my_flow.py on the server, then the following error appears in the logs:

6 August 2021,10:31:49 	agent01	INFO	Submitted for execution: PID: 135

26 August 2021,10:31:49 	agent03	INFO	Submitted for execution: PID: 140

26 August 2021,10:31:49 	agent02	INFO	Submitted for execution: PID: 136

26 August 2021,10:31:49 	execute flow-run	ERROR	Failed to load and execute Flow's environment: ModuleNotFoundError("No module named '/Users/user1/'")

26 August 2021,10:31:49 	execute flow-run	ERROR	Failed to load and execute Flow's environment: ModuleNotFoundError("No module named '/Users/user1/'")

26 August 2021,10:31:49 	execute flow-run	ERROR	Failed to load and execute Flow's environment: ModuleNotFoundError("No module named '/Users/user1/'")

tell me where to look for the error