Core UI - how do I stop a running flow? I couldn’t find any reference to it, and the only way for me to do this right now is to hard kill the agent itself. Any way of doing so from the UI / CLI?

Hi all, is there a recommended approach for objects that cannot be pickled? For example in a machine learning pipeline where you need to do some ETL. Some of the steps require tools that have objects that cannot be pickled so I imagine I might put all those items in one step (SFrames, Spark, etc). Then pipe the results to a CSV file, and the next step we can train a model using the file location of the CSV file, but then I may want one more step to create metrics and send the metrics to MLFlow, but it may be possible that the model cannot be pickled easily, is prefect the wrong tool for this job? But I also would like a cronjob and a UI to look at the results. It seems that I may need to engineer the problem around Prefect vs a single script and Linux crontab. Any suggestions?

I have the same task running at different node of my graph. Now I have a task that must run after one of them is finish. How can I setup this dependency. I am using the Flow framework

I want to get my Parameter at state_handler, It is OK on local test. But it have error on prefect server like this:

Exception raised while calling state handlers: KeyError(<Parameter: fisrt_num>,)

how can i solve this problem? Here is my code

@task
def add(x, y):
    return x+y

def post_to_slack(task, old_state, new_state):
    if new_state.is_failed():
        msg = "Task {0} finished in state {1}".format(task, new_state)
        
    if new_state.is_finished():
        p = new_state.result[num].result
        print(p)
        
    return new_state

with Flow("Tutorial", state_handlers=[post_to_slack]) as flow:
    x = Parameter("fisrt_num")
    y = Parameter("Second_num")

    num = add(x,y)

A kind of a general question to the dev team: are there any plans to allow more granular control over flow execution rights? For example, some of the flows I made are exclusively run automatically by a script (through

client.create_flow_run(_version_group_id_=...)

and there's a chance to screw up some important data if any user tries to run it manually. Another use case I can think of is letting certain members of a team access only specific projects. Thanks!

An unrelated technical question: Currently I'm running DockerAgent on an AWS EC2 server, and it's pulling flow images from an ECR. On that EC2, we need to run

$(aws ecr get-login --region redacted-region --registry-ids 1234567890 --no-include-email)

in order to generate an ECR docker authorization token, and login -- this is standard practice. However! This token is only valid for 12 hours, so the command above is re-run automatically. Now the weird thing is, Prefect Agent doesn't actually "refresh" the auth token and gives me an "Your authorization token has expired" error, even though the command above was run. The only thing that helps is re-starting the Agent, at which point it starts using the new token again. This seems to be some king of a bug, since the configuration of the docker server should not concern the Agent -- or am I missing out on some configuration? Thanks

Hi, quick question, Is it fine to use

if

inside flow like this:

def flow_func(ip):
    with Flow('flow func') as flow:
        task1()
        if ip == True:
            task2()

Hey, a question about secrets. The documentation appears to suggest that setting & retrieving secrets is disabled on a self-hosted server? I appreciated the

cloud

tags throughout the documentation, but I think you’re missing one on the linked header.

Hi all! I'm looking into a single tool to unify data flows and Prefect looks interesting! From the installation docs I can see it's simply a

pip install

. I am curious what high-availability options there are in case the master/scheduling/coordinator(?) node fails, and where the configurations/flows are stored that one can restore from.

I've couple of questions would like to know how will perfect addresses these. • Ability to add human workflow. Basically want to inject an approval/reject or some data entry operations by human. Is it possible to achieve that? • Is it possible to use Apache Arrow to pass data between tasks? Apache Arrow seems to be an interesting data reprasentation where by we can avoid extra serialization/deserialization

is the only way to register a flow by calling the register method on the flow on the same server prefect is running on?

Testing out Prefect as an alternative for an enterprise deployment and running into an odd issue with scheduled flows. For some reason I can only get

IntervalSchedule

tasks to run 10 times (this is using Core/not on cloud). Am I missing something obvious? When I initially register the flow, 10 runs get queued and executed, but no more. In the terminal running the server, the scheduler wakes, schedules 0 runs, and sleeps even if all 10 original runs have passed. On-demand runs in the UI still work at this point. I'll comment with what I'm doing to reproduce.

I see the perfect tasks are not long running. Eg: the database fetch task it opens connection executes query and closes connection. Is it possible to develop a task which runs forever. I can write a PostgreSQL Task which initializes connection pool and it's open forever. All subsequent query executions will re-use connection from pool. Another use case is a gRPC or Arrow Flight IPC server which runs forever and will serve other tasks the necessary data. Something similar to zeebe or netflix conductor provides.

Dear Prefect Community 👋 I’d like to make a Kubernetes secret (containing GCP credentials) available in the worker pods created by the Kubernetes agent. Is this possible?

I have a question about securing Prefect. Right now I have a prefect server deployed to an aws instance. I expose port 8080 to grant access to the server, but it's also required that the graphql port 4200 is also exposed due to the client side javascript. However I don't want graphql to be exposed since that allows the user to send arbitrary commands to the prefect server. Is there anyway to just give access to the web UI, but not allow access to graphql?

What is necessary to get Prefect to use checkpointing? It seems there is something missing from my simplified test code, but unclear what:

from prefect import Flow, task, unmapped, Parameter
from prefect.engine.results import LocalResult
from prefect.engine.executors import LocalDaskExecutor
import prefect

lr = LocalResult(location='{flow_name}-{task_name}-{map_index}.txt')

@task(log_stdout=True, checkpoint=True,)
def add(x, y):
    return x + y

with Flow('iterated map', result=lr) as flow:
    y = unmapped(Parameter('y', default=10))
    mapped_result = add.map([1, 2, 3], y=y)

flow.run(executor=LocalDaskExecutor())

One more question around checkpointing:

target (str, optional): location to check for task Result. If a result
            exists at that location then the task run will enter a cached state.
            `target` strings can be templated formatting strings which will be
            formatted at runtime with values from `prefect.context`

If I have a mapped task, and I want it to create separate outputs per map, I see

map_index

in

prefect.context

, but that relies on my arguments being in the same order each time, correct? Is there a way to pass Parameter values into this

target

or the result's location kwarg such that when I look at the directory containing my checkpoint files, I quickly know which ones have completed?

Hi I'm trying to do a lot of work with mapped functions. How do i merge the results of 2 tasks together? For example

imputed_categorical_dfs

returns two objects. I need to merge these with another task which does the same.

My code which does not work.
transformed_df_map = merge_transformed_data.map(
        df1=[*imputed_categorical_dfs, *encoded_df_map],
        df2=[*imputed_numeric_dfs] + [*yeo_johnson_dfs],
    )

Hiya, for prefect server. Are there any statsd metrics that can be hooked into? Or any events we can use for monitoring prefect internally.

Hey there, I have a problem with prefect server. My setup used to work but now there might be something wrong with the version or something. I’m getting this error in the UI when running a flow:

[3:18pm]: Exception raised while calling state handlers: HTTPError('400 Client Error: Bad Request for url: <http://localhost:4200/graphql/alpha>')

Any idea on how to start debugging?

Hello guys, first time here and recent user of prefect. First of all I need to say it that so far I’m loving it, keep with the awesome work!! I need help to compose a flow with several task with map. Currently I can do it, but each task is only processed after all task from last step are done. This is sub-optimal since each branch of the pipeline(each .map()) are independent of each other. There is any way of achieving that, out-of-the-box??

def run_flow_step_0_map(self, df):        
        cols = list(df.items())
        with Flow("pipeline_map") as flow:
            col = self.get_data.map(unmapped(self), cols)
            col = self.task1.map(unmapped(self), col)
            col = self.task2.map(unmapped(self), col)
            col = self.task3.map(unmapped(self), col)
            col = self.task4.map(unmapped(self), col)
            col = self.task5.map(unmapped(self), col)
            result = self.task5.map(unmapped(self), col)

        return flow

(is confusing the explanation, if you need extra info just say it)

Hi! I'm looking at the code and can't find my way around, can you kindly point me to the place where Flows are run on the server ? I use eliot for logs, and I'd like flow runs to live in an action context. Easily done locally overriding the run method, but I can't find my way around server code ( coming from airflow, love your job, it's a lifesaver ! )

Hello everyone! I just joined this group, as I was assigned to do a POC with Prefect 🙂 I hope I'll get useful information reading through these channels. Cheers!

Hi folks! I'm running prefect on a vm and having issues when the box reboots. The docker containers and prefect server restart just fine, but the agent doesn't--so scheduled jobs don't start. What's the best way to deal with this?

Hi everyone! We’re trialling Prefect (so far good docs and user experience), I’m just overwhelmed when going from a single local environment to a multi-environment distributed setup (this will help a lot https://github.com/PrefectHQ/prefect/issues/2508#issue-613607085 !). For parametrising resources, I tried the below

config.toml

to have different buckets per env:

source_data_bucket = "${environments.${environment}.source_data_bucket}"
transient_data_bucket = "${environments.${environment}.transient_data_bucket}"

[environments]

    [environments.dev-nelson]
        source_data_bucket = "<s3://REDACTED>"
        transient_data_bucket = "<s3://REDACTED>"

    [environments.prod]
        source_data_bucket = "<s3://REDACTED>"
        transient_data_bucket = "<s3://REDACTED>"

I can print the

prefect.config.transient_data_bucket

inside a task, but when used as

S3Result(bucket=prefect.config.transient_data_bucket)

it fails with

Invalid bucket name ""

. How are others doing this? Note I’m providing this result as a task config

@task(
    target="{date:%Y-%m-%d}/crunchbase-raw.cloudpickle",
    result=S3Result(bucket=prefect.config.transient_data_bucket)

hey everyone - I need to connect to a postgres DB through a SSH tunnel. Is there a native way to do this with the

Postgres

task or do I need to make something custom

Hey guys, having 2 problems trying to use FargateTaskEnviornment with FargateAgent. The docs page confused me ever so slightly, so I’m not 100% sure I have it all configured correctly. flow.environment is configured as per the following snippet, and builds and registers with my server:

flow.environment = FargateTaskEnvironment(
        launch_type="FARGATE",
        region="eu-west-1",
        cpu="256",
        memory="512",
        networkConfiguration={
            "awsvpcConfiguration": {
                "assignPublicIp": "ENABLED",
                "subnets": ["subnet-X"],
                "securityGroups": ["sg-Y"],
            }
        },
        family="my_flow",
        taskRoleArn="arn:aws:iam::X:role/Role",
        executionRoleArn="arn:aws:iam::X:role/Role",
        containerDefinitions={
            "name": "my-flow",
            "image": "my-flow",
            "command": [],
            "environment": [],
            "essential": True,
        }
    )

Fargate Agent:

echo "export REGION_NAME=${CDK_DEPLOY_REGION}" > .env
echo "export AWS_DEFAULT_REGION=${CDK_DEPLOY_REGION}" >> .env
echo "export AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID}" >> .env
echo "export AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY}" >> .env
source .env
nohup prefect agent start fargate > prefect_fargate.log 2>&1 & disown

Everything seems fine until I try to run it. There’s no update in the agent logs to show it’s been picked up, and the UI just says:

Failed to load and execute Flow's environment: AttributeError("'str' object has no attribute 'get'")

Anything immediate/obvious/stupid stand out? As an aside I’m guessing I’m missing a bunch of config for the agent, but the docs are a little unclear on that one, the top level of the docs [https://docs.prefect.io/orchestration/agents/fargate.html#installation] seem to suggest you only need the handful of vars I added UPDATE: Some progress, it’s now giving a meaningful error,

Parameter validation failed: Missing required parameter in containerDefinitions[0].logConfiguration: "logDriver

Fixing it

Hey everyone _ I was wondering if someone can tell me why the following code runs fine without errors when calling

flow.run()

but

flow.register()

flags an error mainly the healthcheck throws this error - which I see referenced in the documentation … I am using prefect

v0.11.4

ValueError: Mapped tasks with custom result locations must include
{filename} as a template in their location.

Please see the code below

import os
from prefect import task, Flow
from prefect.engine.results import S3Result


@task
def gen_list():
    return [x for x in range(10)]


@task
def add(x, y):
    return x + y


@task
def multiply(x, y):
    return x * y


result = S3Result(
    bucket=os.environ["AWS_BUCKET"],
    location='prefect-testing/{task_name}/{map_index}.prefect'
)

with Flow('Test Flow', result=result) as flow:
    x = gen_list()
    y = gen_list()
    added = add.map(x, y)
    multiplied = multiply.map(added, added)

# flow runs fine if running locally
flow.run()


"""
flow fails to register
ValueError: Mapped tasks with custom result locations must include
{filename} as a template in their location.
"""
flow.register('Test Project')

Hey guys, having trouble with running a flow on FargateTaskEnvironment. Configuration for the Environment is below..

flow.environment = FargateTaskEnvironment(
        launch_type="FARGATE",
        region="eu-west-1",
        cpu="256",
        memory="512",
        networkConfiguration={
            "awsvpcConfiguration": {
                "assignPublicIp": "ENABLED",
                "subnets": ["subnet-X"],
                "securityGroups": ["sg-Y"],
            }
        },
        family="my_flow",
        taskRoleArn="arn:aws:iam::X:role/CommonSuperRole",
        executionRoleArn="arn:aws:iam::X:role/CommonSuperRole",
        containerDefinitions={
            "name": "my-flow",
            "image": "my-flow",
            "command": [],
            "environment": [],
            "essential": True,
        }
    )

I keep getting this error:

An error occurred (ClientException) when calling the RegisterTaskDefinition operation: Fargate requires task definition to have execution role ARN to support log driver awslogs

So my questions : • I’m using the same Uber Role for both the taskRoleArn and the executionRoleArn, probably not best practice but should work? • I’ve thrown every possible log and cloudwatch related policy/permission at it that I can think of, but nothing is taking. I can provide a dump of the permissions if need be? Any help massively appreciated, fairly stumped on it. Is there any way to get more debug info out of it?