A couple of UI issues here using Prefect 2.0.3: 1. Click a flow run --> if it has a long parameter it goes off the page (image1) 2. Not sure how to reproduce this one, the profile menu sometimes likes to move from where it supposed to be (image2) 3. On the work queue page if you upcoming flow runs that have tags that are too long to display it is suppose to show a "+1". On some of the flows this indicator gets cut off (image3) 4. Why do the "Flow Version" and "Tags" not appear on the flow run right away? As far as I can tell you have to wait until the flow run goes into terminating states to see it change. Can we just get them to show up right away? (image4) 5. If you have multiple flow run the tags order changes for some reason. Create two flows each with a tag of "name", "type" --> the second flow run always swaps the order of this tags for some reason.

👋 Hello, team! Is there any guide that I can run orion server from the source code ?

Because I want to add visualization components to it to enable user create flows by drag/drop in websites.

Hi, I'm using prefect UI on my local machine. the work queues are picking up the flow runs, but there is nothing shown in the queue's Upcoming runs (even when there are runs scheduled in the future). What's the possible reasons?

Hi! I am trying out Orion. The flow framework looks really awesome! But the UI still looks a bit behind Prefect1.0. Or at least, it seems to me so. Is it possible to create deployment from UI? Will you add this feature in future? And what is a roadmap of UI development?

Howdy. We are trying out the beta prefect. We think we have an agent connected but I can’t find anything in the beta site to indicated what agents are currently connected. How do I see that in the new interface?

Hello Team. I have a question regarding Radar. I'm creating a bunch of pairs of "get_x" and "sum_task" tasks each "sum_task" is dependent on a "get_x" instance. Everything is running as expected, but the strange thing is on the Radar: • I would expect every "get_x" task on the same level or circle • then every "sum_task" on their own level or circle. But what I'm getting is: • "get_x" tasks are on random levels. I would understand this, but next read point • only first "get_x" run appears with no dependency, later "get_x" executions appear as if they depend on "sum_task" tasks I'm attaching a screenshot of the Radar UI rendered (Funny Run name "perfect-chicken"). Here's the code I'm running:

from prefect import task, flow, get_run_logger

@task
def get_x(x: int=1):
    return x

@task
def sum_task(x: int=1, y: int=2):
    return x + y

@flow
def flow_exec(init: int=1, end: int=4):
    
    i=1
    x1 = get_x(i)
    sum1 = sum_task(x1)

    i = i+1
    x2=get_x(i)
    sum2 = sum_task(x2)
        
    i = i+1
    x3=get_x(i)
    sum3 = sum_task(x3)
    
    i = i+1
    x4=get_x(i)
    sum4 = sum_task(x4)

flow_exec()

I've coded in this ways (Xn variables) instead of appending variables to a list for simplifying the code. This is just a simple equivalent recreation of a real use case that I have that consists of iterating through a list of integrator.io flows and running them from Prefect instead of using the integrated scheduler on integrator.io I'm having this same issue on my real case. I'm putting this here to see if you can help me understand this Radar UI, maybe I'm just not getting the whole idea, or maybe I can help troubleshooting or providing ideas. Sorry for the long text, I just wanted to be very explicit.

Hello all. I have trouble when deploy prefect orion in kubernetes. I'm using helm to deploy the stack with ingress enabled. I use ingress path domain.mycompany.com/orion to orion service, but the UI still request static files (.js, .css) to domain.mycompany/<static-files> and not to domain.mycompany.com/orion/<static-files> , is that something wrong??

On prefect cloud 2.0, my deployment is showing up but I am unable to click Run?

Hi folks, in the prefect 2 UI, is there a way I can view a specific task run’s logs ? i.e. isolate the logs to only a particular task run

Hello, I am trying to use the prefect 2 UI. It would be nice to be able to see the flows that I am executing in the browser. I must do something wrong or maybe my understanding is wrong. Won"t the flows that I run on the commandline then show up in the flows page in orion?

I am not on a laptop but doing this on a remote server, so I am starting orion like this:

PREFECT_ORION_API_HOST=0.0.0.0 prefect orion start

Hello, one question regarding Storage. In the documentation it says

If no storage is explicitly configured, Prefect will use

LocalFileSystem

storage by default. Local storage works fine for many local flow run scenarios, especially when testing and getting started. However, due to the inherit lack of portability, many use cases are better served by using remote storage such as S3 or Google Cloud Storage.

If I add

--storage-block s3/example-block

when doing

prefect deployment build

I get a copy of all my files on my S3 bucket. But how can I ensure that the Deployment in Prefect Cloud always takes the flow codes from the S3 bucket?

Hi together, I'm trying out the Slack notification example and have a question about the Result object. Is it possible to display the error message or exception?

Hi all I have a task which a series of string inputs:

competitor

,

lang_code

,

model

and

work_dir

@task(name="test_eval", log_stdout=True)

def run_test(
    competitor: str, 
    lang_code: str,
    testset_config: dict,
    model: str, 
    work_dir: str
    
    ):
    testset_path = testset_config[lang_code]
    savepath = f"/exp/aniruddhas/comp_eval/test_outputs/2022-08-16/test_dynamic_outputs/{competitor}_{testset_path}_{lang_code}_{model}.txt"

    return f'aladdin/evaluation/test.sh --competitor {competitor} --testset_path "{testset_path}" --lang_code {lang_code} --model "{model}" --work_dir "{work_dir}" | tee {savepath}'

From here, I have a list of competitors and a list of languages. I want to be able to create a Shell Prefect task but mapped over a list of competitors and languages. So far I have this:

shell_result = shell_task(
            command=run_test.map(
                competitors, 
                languages, 
                testset_config=unmapped(testset_config), 
                model=unmapped(model), 
                work_dir=unmapped(work_dir)
                )
            )

But this doesnt seem to work. I want a ShellTask to be able to iterate through each language for each competitor. How can I write this?

Hi, I am still trying to get some output in the orion web gui. It is always empty for me and I am trying for several days now. I have 2 docker containers and they seem to talk to each other just fine:

sorry for pasting cli output as an image btw. any idea why I can"t get any output from orion?

I believe something should appear by itself, but I click through all menus: Flow runs, Flows, Deploments, Work Queues, Blocks, Notifications. There is an empty graph in Flow Runs. All other menus show some 'create' or 'add' empty state div

I used this image: prefecthq/prefect:2.0.4-python3.7 - see its output on the right in the foto above

In the left pane in the picture I executed this from a different python 3.7 container in the same docker network:

from prefect import flow
    @flow(name="Prefect 2.0 Flow")
    def flow_prefect():
        print("Hello Prefect 2.0!")
        import time
        time.sleep(5)
        print("Good Bye Prefect 2.0!")

    flow_prefect()

I found this and I am not sure if it is related. It says the UI makes request to localhost, and that localhost is hardcoded https://github.com/PrefectHQ/prefect/issues/5648

There is a recommendation in the github issue: set

PREFECT_ORION_API_HOST="127.0.0.1"

and start the server with

prefect orion start --host 0.0.0.0

After some more investigation: it really helped to use a different docker container for orion and for the python flows. It is good to have the configs for client and server separated. Trying out different combinations of settings for the IP configuration of the server.

OK, it is solved. I will update later, need a break now.

Hi, apologies if this is not the dedicated channel for this type of question. I am trying to deploy a flow to a self remote hosted Orion server (docker, Prefect V2) using S3 block as a code repository.

prefect deployment build ./flows/my_awesome_flow.py:my_stars -n flow1 -t dev -sb s3/s3test

I got the following error:

RuntimeError: File system created with scheme 's3' from base path '<s3://my_s3_bucket>' could not be created. You are likely missing a Python module required to use the given storage protocol.

the error does not have enough info what module I miss. any idea is appreciated ?

Hello, For Prefect 1.0 there was documentation with some resources for the different kind of remote Agents: Docker, ECS, ... For Prefect 2.0 I do not see here if there are different alternatives for remote agents. Is there any tutorial or similar where is shown how to launch a remote agent? Preferably if it is listening to a Deployment scheduled in Prefect Cloud. Thanks in advance!

Hello. I'm getting started with Prefect v2 and I'm having an error with setting up the Kubernetes Job with Prefect Cloud. I'm trying to configure the Customizations field to modify the output manifest. However, if I write it that way, there is this error while I run the prefect CLI:

$ prefect deployment build ./expenses.py:export_raw_expenses -n export-raw-expenses -t kubernetes -ib kubernetes-job/job --storage-block azure/azure-storage

Found flow 'export-raw-expenses'
Successfully uploaded 2 files to <az://prefect/htoh-prefect>
Traceback (most recent call last):
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/cli/_utilities.py", line 41, in wrapper
    return fn(*args, **kwargs)
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/utilities/asyncutils.py", line 193, in wrapper
    return run_async_in_new_loop(async_fn, *args, **kwargs)
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/utilities/asyncutils.py", line 140, in run_async_in_new_loop
    return anyio.run(partial(__fn, *args, **kwargs))
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/anyio/_core/_eventloop.py", line 70, in run
    return asynclib.run(func, *args, **backend_options)
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 292, in run
    return native_run(wrapper(), debug=debug)
  File "/usr/local/Cellar/python@3.10/3.10.6_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/asyncio/runners.py", line 44, in run
    return loop.run_until_complete(main)
  File "/usr/local/Cellar/python@3.10/3.10.6_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/asyncio/base_events.py", line 646, in run_until_complete
    return future.result()
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 287, in wrapper
    return await func(*args)
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/cli/deployment.py", line 477, in build
    infrastructure = await Block.load(infra_block)
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/blocks/core.py", line 514, in load
    return cls._from_block_document(block_document)
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/blocks/core.py", line 415, in _from_block_document
    block = block_cls.parse_obj(block_document.data)
  File "pydantic/main.py", line 521, in pydantic.main.BaseModel.parse_obj
  File "/Users/manhha/Library/Caches/pypoetry/virtualenvs/htoh-prefect-8QVbx3Ri-py3.10/lib/python3.10/site-packages/prefect/blocks/core.py", line 103, in __init__
    super().__init__(*args, **kwargs)
  File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__
pydantic.error_wrappers.ValidationError: 1 validation error for KubernetesJob
customizations
  string indices must be integers (type=type_error)

It looks like the string that the system doesn't like the string I input to the Customizations field:

{
  "op": "add",
  "path": "/spec/template/spec/containers/0/imagePullPolicy",
  "value": "Always"
}

I tried to consult the doc but it only mention the JSON 6902 patch, but nothing more. There is however an example in the Github repo of Prefect. Do you know if I'm missing something?

Is there a possibility or some instructions for installing Prefect 2 UI on Kubernetes?