Hi prefect - I've set up a teams webhook automation to alert me when queues have become unhealthy queue-name unhealthy! Name: queue-name Last polled: None Late run count: URL: xxxx Using the out-of-the-box automation. The queue in the UI is displaying as healthy, but for some reason does not have a last polled value, hence it keeps displaying as unhealthy. Any suggestions here?

Hey all - I find the deployments screen confusing - would love to be able to sort using the table headers but even better group by the 'project' tags I have specified or similar. I know I can sort a-z using the dropdown but I think in a large project the default sort should be a-z rather than created

Hello! I recently set up a new automation in prefect 2 cloud, but having issues viewing the events. When I click on events I can see the list of events from the automation, but when I click into an individual event I am getting a blank page. I also noticed in the network traffic that I get a

403 forbidden

when it filters for an event in the UI

Hi, I am trying to get started with prefect cloud. I have a server that is supposed to run the jobs. I logged that server into the prefect cloud and started a task-runner. I have different flows and a few deployments. I could run all of these locally. I am now seeming to run into path issues. The deployments are created using Deployment.build_from_flow where path is

path=os.getcwd() + "/flows

and

entrypoint="flow.py:test_flow

The test flow runs the following command:

shell_run_command(
        command="python3 src/task.py", return_all=True
)

The project structure is as follows: project - flows -- depolyments.py -- flow.py - src -- task.py I run the task runner from the project root with

prefect agent start -p default-agent-pool

I applied the deployments and see them in the cloud and can start them from there. The error I get is:

can't open file /tmp/tmp6ohfe_wcprefect/src/task.py [Errno 2] No such file or directory

After exiting the process not even

/tmp/tmp6ohfe_wcprefect

exists, but I dont know if this is because prefect tears it down. Any help regarding this would be very appreciated. I hope i provided enough information to help with this. If not, let me know.

Hi dear Prefect team. I am having issues with the UI and my flows. I am creating some flows that spawn 4-8 tasks at a time and are also chained with some other as well (So not many tasks). The issue is that when I want to see the flow in the cloud UI, it does not even render the timeline or the radar view. It can take hours and it slows down the web browser without any actual result (It looks like the attached image). I thought that it had to do with how heavy/complex are the tasks but it has nothing to do. The attached example is a flow that is only printing configurations and sleeping for 2 seconds (So this rendering issue also happens with simple tasks)

dear Prefect team, I am using prefect v1 on the cloud. When I push a flow that build with docker and KubernetesRun through ci/cd, the schedule mode of this flow is automatically turned on, while I want it to stay off. How to avoid that ?

Hi, So i know in the work pools you can limit concurrency to only one. But is there a way to set up an agent or work pool or the deployment that limits the deployment to a concurrency of one. Because currently how i'm doing is sending the deployment to it's own work pool that has a limit concurrency of one but then i have a bunch of work pools. Now i know there's a task run concurrency but i don't want to just limit a singular task i want to limit the whole deployment. So basically asking if there's a way to stop the deployment from running if only it currently is already running. Regardless of other deployment flows Thanks,

Hi Prefect Team, I am running into an issue with the Prefect 2.8.4 UI where I can’t access details for any of my deployments. (i.e. the

/deployments/deployment/<uuid>

route in the UI). I’ve included a screenshot, my parameters are really simple, just a single level Pydantic (v 1.10.4) object with some floats. Do you have any advice on how I could debug this?

Hi Dejapong, I've reported the same issue in the prefect-community.

I've seen that if you do in the three dots of any deployment Quick run or Custom run it works. But we are not able to see the deployment

Unable to spin docker-compose with Prefect 2.8.4/older official docker-hub images with Postgres: I would love if someone can point out why this is failing. Here's the error (I had to trim due to character limitations) -

2023-03-06 20:19:39     value = await result
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/connection.py", line 2092, in connect
2023-03-06 20:19:39     return await connect_utils._connect(
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/connect_utils.py", line 895, in _connect
2023-03-06 20:19:39     raise last_error
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/connect_utils.py", line 881, in _connect
2023-03-06 20:19:39     return await _connect_addr(
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/connect_utils.py", line 773, in _connect_addr
2023-03-06 20:19:39     return await __connect_addr(params, timeout, True, *args)
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/connect_utils.py", line 825, in __connect_addr
2023-03-06 20:19:39     tr, pr = await compat.wait_for(connector, timeout=timeout)
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/compat.py", line 56, in wait_for
2023-03-06 20:19:39     return await asyncio.wait_for(fut, timeout)
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/asyncio/tasks.py", line 445, in wait_for
2023-03-06 20:19:39     return fut.result()
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/site-packages/asyncpg/connect_utils.py", line 684, in _create_ssl_connection
2023-03-06 20:19:39     tr, pr = await loop.create_connection(
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/asyncio/base_events.py", line 1036, in create_connection
2023-03-06 20:19:39     infos = await self._ensure_resolved(
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/asyncio/base_events.py", line 1418, in _ensure_resolved
2023-03-06 20:19:39     return await loop.getaddrinfo(host, port, family=family, type=type,
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/asyncio/base_events.py", line 863, in getaddrinfo
2023-03-06 20:19:39     return await self.run_in_executor(
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/concurrent/futures/thread.py", line 58, in run
2023-03-06 20:19:39     result = self.fn(*self.args, **self.kwargs)
2023-03-06 20:19:39   File "/usr/local/lib/python3.10/socket.py", line 955, in getaddrinfo
2023-03-06 20:19:39     for res in _socket.getaddrinfo(host, port, family, type, proto, flags):
2023-03-06 20:19:39 socket.gaierror: [Errno -2] Name or service not known
2023-03-06 20:19:39 
2023-03-06 20:19:39 Application startup failed. Exiting.

Here's my docker-compose -

version: '3.8'

networks:
  prefect:
    name: prefect

services:
  # ------------------------------------------------------------- #
  #             PostgreSQL DB for Prefect DB                      #
  # ------------------------------------------------------------- #
  postgres:
    image: postgres:latest
    restart: always
    command:
      - postgres
      - -c
      - max_connections=150
    environment:
      - POSTGRES_USER=${POSTGRES_USER}
      - POSTGRES_PASSWORD=${POSTGRES_PASSWORD}
      - POSTGRES_DB=${POSTGRES_DB}
    healthcheck:
      interval: 10s
      retries: 60
      start_period: 2s
      test: pg_isready -q -d $${POSTGRES_DB} -U $${POSTGRES_USER} | grep "accepting connections" || exit 1
      timeout: 2s
    ports:
      - 5432:5432
    expose:
      - 5432
  # -------------------------------------- #
  #             Prefect Server              #
  # -------------------------------------- #
  prefect-server:
    image: prefecthq/prefect:2.8.4-python3.10
    command:
      - prefect
      - server
      - start
    ports:
      - 4200:4200
    expose:
      - 4200
    environment:
      PREFECT_API_DATABASE_CONNECTION_URL: postgresql+asyncpg://${POSTGRES_USER}:${POSTGRES_PASSWORD}@postgres:5432/${POSTGRES_DB}
    networks:
      - prefect
    restart: always
    depends_on:
      postgres:
        condition: service_started

Hello all! I wanted to share a quick UI update that we've added the ability to use markdown for your deployment descriptions. You can see it now in cloud and should be able to see it in OSS after our next release.

Hey everyone, I currently don’t see the

AwaitingRetry

state available as a filter in the UI, could it be added there? (I haven’t found any issue related to this topic). Also, queues are not set as

unhealthy

when there are flow runs

AwaitingRetry

, I’m not sure this is the desired behavior…

Hey - The SQLAlchemy Connector block UI seems to be broken - if you click the connection info choices you lose the driver fields and cant get them back

Hey folks! In the next week or so we’ll be releasing an update to the Flow Run Timeline and retiring the Radar graph. Here’s more info about it (video). Would love to hear your thoughts! https://www.loom.com/share/6b5975de4a6e455a9340c7b99cdf5aa4

looks like a great improvement to me and I really like adding more and more interactivity to it, that’s super useful for navigating to a specific task without going through the list of tasks/subflows

unrelated question, is there a way to see flow runs beyond last 7 days in the deployment view?

I don’t see any docs on the new Results tab - are those docs coming? or did i just miss them

Hey everyone! We just released a new Results tab within Flow Runs that allows you to see information about results from a flow or task. As a reminder, by default Prefect tracks only metadata about your results to help you understand how data is passed between your tasks. Here’s more info about the new Results view (video): https://www.loom.com/share/bcf4e08798244790a3383c0e5276c173

Hello! is there a way to change how many flow runs the UI request to the backend by default? I am getting huge payloads which cause the UI to not load any flow runs

Hello @here i am having a issue with retry task status says it is retrying but task is now running and task status is not changing for like 30 minutes. PS : i am running on local server here i have putup some details https://stackoverflow.com/questions/75840750/prefect-task-status-shows-retrying-while-it-srarted-running

Hi prefect team. I saw that the runtime flow visualization has been updated. I definitely like it better. I was curious if there was a plan for a compiled static dag view, similar to the 'Schematic' view from Prefect 1? Seeing a schema for a DAG is hugely valuable for collaboration, e.g. someone is working on a pipeline, they ask me to look at it. I have to currently dig into specific flow runs and look at something like this, to get my 1000 foot view of what this pipeline even is. I'd much rather not see a run time visualization because the time taken might vary a lot, and the edges are still all crossed so I have to click on each task and trace my eyes along the screen to see the dependency. Will add in the comments an example preferred view. Ideally this DAG schematic would just live on the Flow's page on a separate tab, similar to Prefect 1.

This message was deleted.

Any chance, anyone else is having this particular issue?

One thing that I find a little bit counter intuitive is the inability to search by flow name in the deployment panel. I know it's a deployment panel but since the name contains both the flow name and the deployment name, I would expect this to work. Otherwise it can bring the "bad habit" of also adding your flow name to your deployment name and it gets weird.

Hi everyone! I have a question for Prefect `2.8.5`: Is it possible to get unformatted output when viewing flow run logs? I sometimes want to copy/paste a multi-line error to a teammate, but it comes out looking very bad because of the log level/timestamp columns. It would be great to have the option to show them unformatted.

Hey all, The latest release includes more performance enhancements to the Flow Run Graph, particularly for runs of ~5k+. Would love to hear how it’s working out for folks, good or bad. Also includes flow names prepended to sub-flow runs (e.g

my-flow / raging-platypus

)

Just a heads up, I have noticed that the UI is consistently inconsistent when navigating between views. Refreshing the page fixes the issues. Common issues I find: • bars do not load in the run graph • text doesn't stay in their containers • on click events for navigation become unresponsive Attached is a great example of the text container issue. This one did not get resolved with a page refresh. You can see that the log content container stretches almost to the end of the details container.

Hi Team, Can we have 2 prefect versions(1.X and 2.X) in our local machine?

Hi! My deployment previously used dockerinfrastructure, which I updated to process. But the deployment still somehow references my old docker infrastructure, and I am not able to delete the old block. "Failed to delete block"