I'm following this doc - https://docs.prefect.io/orchestration/execution/overview.html#labels

Issue with keeping Agents alive with Supervisor. I am having an issue with properly initialising supervisor to keep my agents running. I have generated the generic supervisor code but I keep getting the following error messages: "can't find command 'prefect'". Any ideas what might be the issue here? This is my supervisor code:

[unix_http_server]
file=/tmp/supervisor.sock

[supervisord]
loglevel=debug

[rpcinterface:supervisor]
supervisor.rpcinterface_factory = supervisor.rpcinterface:make_main_rpcinterface

[supervisorctl]
serverurl=unix:///tmp/supervisor.sock

[program:local-agent]
command=prefect agent local start -l local

[program:docker-agent]
command=prefect agent docker start -l docker

The main issue I am trying to solve is keeping the agent up and running at all times. My VM does a scheduled restart once a day which means any agents I spin up manually will be shut down. Is there any way of setting the agent itself up as a docker container which starts on startup like the prefect server itself? It seems like the docker agent itself isnt actually running in docker but rather just issues the flow to its own docker container?

Hey guys, I've upgraded from prefect orion

2.0a9

to

2.0a12

. When I try run the server

prefect orion start

it throws this error:

sqlalchemy.exc.OperationalError: (sqlite3.OperationalError) table flow already exists

how do I get around this?

I am trying to follow the intro to sub-flow results passing and for some reason, the get_task_run_results task get stack. I am using prefect server and have kubernetes agent (though I am not sure that matter since I use the --execute cli arg.

I have a flow A with multiple sequential tasks. When the last task of Flow A completes, I want to kick off flow B, could I do something like :

with Flow("FlowB") as flow:
    task_c = do_task_c()
    # ...
    

with Flow("FlowA") as flow:
    task_a = do_task_a()
    task_b = do_task_b(upstream_tasks=[task_a])
    # ... 
    create_flow_run(flow_name="FlowB")

though just want to confirm whether if I get create_flow_run to wait for task b to complete? (what would be the best way to arrange this?)

Hi, we are using prefect together with dask for scheduling ML training runs. We use a temporary cluster. For big grid searches we see that the dask workers accumulate memory over time and then stop accepting tasks. I assume similar to this thread. However none of the recommended solutions helped. Has anyone had similar experience? Could this be related to prefect?

Is there any way to set a schedule as "inactive" on registration? Every time our CI/CD system rebuilds and registers, any schedules that we have set as inactive in the UI get reactivated. This is particularly an issue for our dev environment, where we don't usually want flows running on a schedule. Thanks!

hi in my prefect flow running on prefect cloud am setting the current directory as the directory where my prefect flow available , but the other python modules which is called from the prefect flow is in different directory , this is actually throwing error when called from prefect cloud , error says that particular python module not found since it is another directory if I copy my prefect flow file to the pth where python modules available then no error so is there anyway I can keep my python modules in one folder and flow in different folder

hi all, I'm trying to do a conditional flow, but as you can see i get matches, check if the matches are valid, and then only want to continue if they are. Because i am passing matches into those tasks, they run regardless of if the condition is true or not. What am i doing wrong here?

with flow:
    start_date, end_date = Parameter("start-date", default=None), Parameter("end-date", default=None)
    start_date, end_date = set_date_params(start_date, end_date)
    matches = get_completed_matches_by_date_range(start_date, end_date)
    there_are_matches = are_there_any_matches(matches)
    with case(there_are_matches, True):
        some_data = some_task.run(matches)

Hello! Is it possible to retrieve the flow run ID from the

Flow

or

State

objects? I'm trying to get it using the

on_failure

callback for my flow

Hey guys, with regards to retrying tasks on specific exceptions, is the current solution still this? https://github.com/PrefectHQ/prefect/discussions/4956 (similar question asked here)

hey community, I have a question that I've been unable to find an answer for in the docs. say I have two mutually exclusive flows, Flow A and Flow B. Flow B should not be executed until Flow A has completed successfully. how can I declare a dependency on Flow A from within Flow B? dagster has something similar to this and I was hoping that prefect does as well.

Hey all! I'm trying to use

DocekrStorage

, and my

Dockerfile

looks like the below. It gets through all the steps except the final step with an error. Does anyone know how to resolve this?

Hi there,

I'm using prefect orion but I can't find tasks libraries.

(from prefect.tasks.notifications import email)

seems not working in orion. am I wrong ?

Join us in 2 mins to learn how to generate end-to-end lineage graphs of your data pipelines using Monte Carlo and Prefect so you can reduce the time to detection and resolution for critical data incidents: https://www.montecarlodata.com/how-to-build-more-reliable-data-pipelines-with-monte-carlo-and-prefect/

Hi does anyone have an example on using the extra loggers? I have a snippet in

custom.py

like this

import logging
import sys

def run():
  logger = logging.getLogger("custom")
  logger.setLevel('INFO')
  log_stream = logging.StreamHandler(sys.stdout)
  log_stream.setFormatter(LOG_FORMAT)
  logger.addHandler(log_stream)

and also have

[logging]
# Extra loggers for Prefect log configuration
extra_loggers = "['custom']"

in the toml config, but logs from the custom logger don’t seem to be captured by Prefect and show up in the cloud console.

Alright, I'm slowly pushing further with

DockerStorage

and then realized I can't use a

local

agent to run flows that use that. No problem... since my local agent is setup in a docker container, I figured I would just use a

DockerRun

config. This seems to go alright at the very first, but then it can't pull the image from our ECR. 🧵

Hi everyone, I’m having some trouble getting an ECS run to work with a custom docker image. Whenever I run the flow it gets stuck in the “submitted for execution” state. I’ve checked the CloudWatch logs but it only shows that the flow deployment was completed. Does anybody have any suggestions for how to resolve this problem? My suspicions are that the custom image is unable to be pulled from the private ECR repo, so does anybody know what the correct permissions I need to access ECR? Below is my RUN_CONFIG for ECSrun:

RUN_CONFIG = ECSRun(
    labels=["dev"],
    task_role_arn=f"arn:aws:iam::xxx:role/prefectTaskRole",  # a role with S3 permissions
    execution_role_arn="arn:aws:iam::xxx:role/prefectECSAgentTaskExecutionRole",
    run_task_kwargs=dict(cluster="prefectEcsClusterDev"),
    image="<http://xxx.dkr.ecr.us-east-1.amazonaws.com/prefect-orchestration:docker-test-v1|xxx.dkr.ecr.us-east-1.amazonaws.com/prefect-orchestration:docker-test-v1>"
)

Hi everyone! I've got two agents on my k8s cluster, both of them are on the

prefect

namespace and work fine. I needed to add one more agent to the cluster, but this one should be on another namespace, say

prefect-agent-xxxx

. When I do this, I can successfully submit runs to it, they do get deployed, but it doesn't seem to actually run and no logs are shown. I've tried configuring the apollo URL to

<http://prefect-apollo.prefect.svc.cluster.local:4200>

and also setting an

ExternalName

to it in the

prefect-agent-xxxx

namespace and using it, but none of them works. Any ideas on how I could debug this?

I am a little bit confused at the way prefect treats mapping. If you map to an empty list/tuple do no child tasks get created and the task is skipped? Or is are all child tasks failed automatically?

Hi, I had a few issues/questions which I will ask in different posts: post 1)I have setup two different python environments. In the first environment the latest prefect version is installed and jobs run to completion. Using the first environment I create a second environment and install tensorflow 2.4.0 which causes some of the packages such as numpy to be downgraded from 1.22.x to 1.19.5 . In the second environment jobs either hang (stop running) or there is a segmentation fault, depending on the OS. Is there a reason for this behavior? Any pointers I could use or info I could provide? I am using the LocalDaskExecutor for running jobs locally.

post 2) I have also made a distributed environment where I am using the prefect server/agent and dask-schedular/worker. When running the dask worker locally (i.e all in the same machine) jobs seems to run okay however when running a worker on a different machine following is some of the errors that I get (the prefect version is the same across the machines): raise ClientError(result["errors"]) prefect.exceptions.ClientError: [{'message': 'Invalid task ID', 'locations': [{'line': 2, 'column': 5}], 'path': ['get_or_create_task_run_info'], 'extensions': {'code': 'INTERNAL_SERVER_ERROR', 'exception': {'message': 'Invalid task ID'}}}] ERROR - 2022-02-25 130157,769 - task_runner - Failed to retrieve task state with error: ClientError([{'message': 'Invalid task ID', 'locations': [{'line': 2, 'column': 5}], 'path': ['get_or_create_task_run_info'], 'extensions': {'code': 'INTERNAL_SERVER_ERROR', 'exception': {'message': 'Invalid task ID'}}}])

post 3) This is related to a post 1 above, will it be possible to have prefect/dask setup to run jobs on two different python environments, one running tensorflow with perhaps older packages and the other running a newer environment.

Hey, I am running a flow locally and have another flow deployed in the cloud. I loop over something and create multiple sub flows.

create_flow_run.run(
            flow_name="advanced_cleaning",
            project_name="modelling",
            run_name=f"data_update={data_update}",
            parameters=dict(data_update=data_update),
        )

When I run this it is called multiple times, but only one sometimes two flows are triggered in the cloud, is this a possible bug? Do run names need to be unique? This is being called from within a task fyi

Is it possible to run dask code within prefect task? Is there a way to get submit work to the dask cluster that runs my flow (KubeCluster to LocalDaskCluster)? I am looking for a way to make my core functionality rely on dask and still be able to leverage the orchestration through prefect...

Hi I am new to this , I am using python, after initialising a prefect flow , we use flow.run() to start the flow . I would like to know any command which can terminate this flow . Did not find any such thing on prefect doc.

Hi everyone, I have a question regarding Prefect Core ~ Prefect Orion. Is it possible to use Prefect Core’s tasks inside Prefect Orion ? Meaning can I use the DBT Task library for core in Orion.

I have a general questions about updating tasks in a flow: say you deploy a flow with an interval schedule, you realise after a few flow runs that there is a bug in one of the tasks, you make a fix for this, is there a way to update an already deployed flow with this fix or do you need to delete the deployed flow and deploy again with the new code?

I am trying to replicate running this flow on my own laptop with Docker Desktop. I have created a docker image (but did not push it to e.g. DockerHub) and registered the flow just like in the script to Prefect Cloud. Started a docker agent and did a quick run from the UI. So far so good everything works. But, when I modify the Dockerfile so that it can run as non-root, I can no longer do a successful run. I get the following error from the agent:

docker.errors.NotFound: 404 Client Error for <http+docker://localhost/v1.41/containers/f30a3e9241fab0d272c3f36eae16867487fb187b964c8b3f22bc8fd05d2aa4d0/json>: Not Found ("No such container: f30a3e9241fab0d272c3f36eae16867487fb187b964c8b3f22bc8fd05d2aa4d0")

and the flow is stuck in submitted state. Does anyone knows how to fix it? Just to make sure, the only difference between the successful run and the run stuck in submitted state is the fact that I added a non-root user to the image. https://github.com/anna-geller/packaging-prefect-flows/blob/master/flows_no_build/docker_script_docker_run_local_image.py