Skip to content

HuangJiameng/dflow

 
 

Repository files navigation

DFLOW

Dflow is a Python framework for constructing scientific computing workflows (e.g. concurrent learning workflows) employing Argo Workflows as the workflow engine.

For dflow's users (e.g. ML application developers), dflow offers user-friendly functional programming interfaces for building their own workflows. Users need not be concerned with process control, task scheduling, observability and disaster tolerance. Users can track workflow status and handle exceptions by APIs as well as from frontend UI. Thereby users are enabled to concentrate on implementing operations (OPs) and orchestrating workflows.

For dflow's developers, dflow wraps on argo SDK, keeps details of computing and storage resources from users, and provides extension abilities. While argo is a cloud-native workflow engine, dflow uses containers to decouple computing logic and scheduling logic, and uses Kubernetes to make workflows observable, reproducible and robust. Dflow is designed to be based on a distributed, heterogeneous infrastructure. The most common computing resources in scientific computing may be HPC clusters. User can either use executor to manage HPC jobs using DPDispatcher plugin, or use virtual node technique to uniformly manage HPC resources in the framework of Kubernetes (wlm-operator).

OP template (abbr. OP) in dflow can be reused among workflows and shared among users. Dflow provides a cookie cutter recipe dflow-op-cutter for template a new OP package. Start developing an OP package at once from

pip install cookiecutter
cookiecutter https://github.com/deepmodeling/dflow-op-cutter.git

Dflow provides a debug mode for running workflows bare-metally whose backend is implemented in dflow in pure Python, independent of Argo/Kubernetes. The debug mode uses local environment to execute OPs instead of containers. It implements most APIs of the default mode in order to provide an identical user experience. The debug mode offer convenience for debugging or testing without container. For the clusters having problem deploying docker and Kubernetes and difficult to access from outside, the debug mode may also be used for production, despite less robustness and observability.

1. Overview

1.1. Architecture

The dflow consists of a common layer and an interface layer. Interface layer takes various OP templates from users, usually in the form of python classes or functions, and transforms them into base OP templates that common layer can handle.

dflow_architecture

1.2. Basics

1.2.1. Parameters and artifacts

Parameters and artifacts are data stored by the workflow and passed within the workflow. Parameters are saved as text which can be displayed in the UI, while artifacts are saved as files. Parameters are passed to an OP with their values, while artifacts are passed as paths.

1.2.2. OP template

OP template (abbr. OP) is a fundamental building block of a workflow. It defines a particular operation to be executed given the input and the expected output. Both the input and output can be parameters and/or artifacts. The most common OP template is the container OP template. Two types of container OP templates are supported: ShellOPTemplate, PythonScriptOPTemplate. ShellOPTemplate defines an operation by shell script and a container image where the script runs. PythonScriptOPTemplate defines an operation by Python script and a container image.

As a more Python-native category of OP templates, PythonOPTemplate defines OPs in the form of Python classes or Python functions (called class OP or function OP correspondingly). As Python is a weak typed language, we impose strict type checking to Python OPs to alleviate ambiguity and unexpected behaviors.

For an class OP, the structures of the inputs and outputs of an OP are declared in the static methods get_input_sign and get_output_sign. Each of them returns a dictionary mapping from the name of a parameter/artifact to its type. The execution of the OP is defined in the execute method. The types of the parameter values passed in and out should be in accord with those declared in the sign. Type checking is implemented before and after the execute method. For an input/output artifact, its sign should be like Artifact(type) where type can be Path, List[Path], Dict[str, Path] or dflow.python.NestedDict[Path]. For input artifact, the execute method will receive a path, a list of paths or a dictionary of paths according to its sign . OP developer can directly process the file or directory at the path. For output artifact, the execute method should also return a path, a list of paths or a dictionary of paths according to its sign.

from dflow.python import OP, OPIO, OPIOSign, Artifact
from pathlib import Path
import shutil


class SimpleExample(OP):
    def __init__(self):
        pass

    @classmethod
    def get_input_sign(cls):
        return OPIOSign(
            {
                "msg": str,
                "foo": Artifact(Path),
            }
        )

    @classmethod
    def get_output_sign(cls):
        return OPIOSign(
            {
                "msg": str,
                "bar": Artifact(Path),
            }
        )

    @OP.exec_sign_check
    def execute(
        self,
        op_in: OPIO,
    ) -> OPIO:
        shutil.copy(op_in["foo"], "bar.txt")
        out_msg = op_in["msg"]
        op_out = OPIO(
            {
                "msg": out_msg,
                "bar": Path("bar.txt"),
            }
        )
        return op_out

The above example defines an OP SimpleExample. The operation is to copy the input artifact foo to output artifact bar and to copy the input parameter msg to output parameter msg.

For an function OP, the structure of the inputs and outputs are declared in the type annotations more compactly and execution is defined in the function body. Type checking is implemented before and after the function as well. We recommend python>=3.9 to use this syntax sugar. See more about Python Annotation at Python official howtos.

from dflow.python import OP, Artifact
from pathlib import Path
import shutil

@OP.function
def SimpleExample(
		msg: str,
		foo: Artifact(Path),
) -> {"msg": str, "bar": Artifact(Path)}:
    shutil.copy(foo, "bar.txt")
    out_msg = msg
    return {"msg": out_msg, "bar": Path("bar.txt")}

To define an OP template from the above class or function, we need to specify the container image and other optional arguments to PythonOPTemplate. pydflow need not to be installed in this image because local pydflow package will be uploaded into the container by default

from dflow.python import PythonOPTemplate

simple_example_templ = PythonOPTemplate(SimpleExample, image="python:3.8")

An example is here

1.2.3. Step

Step is the central block for formulating rules of data flows. A step is the result of instantiating an OP template, where values of all input parameters and sources of all input artifacts declared in the OP template must be specified here. The input parameters/artifacts of a step may be either static at the time of submission, or dynamically from outputs of another step.

from dflow import Step

simple_example_step = Step(
    name="step0",
    template=simple_example_templ,
    parameters={"msg": "HelloWorld!"},
    artifacts={"inp_art": foo},
)

Note that foo here is an artifact either uploaded from local or output of another step.

1.2.4. Workflow

Workflow connects blocks together to build a workflow. A simple serial workflow is created by adding steps in sequence. Adding a list of steps to a workflow means these steps running in parallel.

from dflow import Workflow

wf = Workflow(name="hello-world")
wf.add(simple_example_step)

Submit a workflow by

wf.submit()

An example is here

2. Quick Start

2.1. Setup Argo Server

If you have an Argo server already, you can skip this step. Otherwise you can follow the installation guide.

2.2. Install dflow

Make sure your Python version is not less than 3.6 and install dflow

pip install pydflow

2.3. Run an example

There are several notebook tutorials that can help you start to use dflow. Besides, you can submit a simple workflow from the terminal

python examples/test_python.py

Then you can check the submitted workflow through argo's UI.

3. User Guide (dflow-doc)

3.1. Common layer

3.1.1. Workflow management

After submitting a workflow by wf.submit(), or getting a history workflow by wf = Workflow(id="xxx"), one can track its real-time status with APIs

  • wf.id: workflow ID in argo
  • wf.query_status(): query workflow status, return "Pending", "Running", "Succeeded", etc.
  • wf.query_step(name=None): query step by name (support for regex), return a list of argo step objects
    • step.phase: phase of a step, "Pending", "Running", Succeeded, etc.
    • step.outputs.parameters: a dictionary of output parameters
    • step.outputs.artifacts: a dictionary of output artifacts

3.1.2. Upload/download artifact

Dflow offers tools for uploading files to the artifact repository and downloading files from it (default artifact repository is Minio set up in the quick start). User can upload a file/directory, a list of files/directories or a dictionary of files/directories and get an artifact object, which can be used as argument of a step

artifact = upload_artifact([path1, path2])
step = Step(
    ...
    artifacts={"foo": artifact}
)

User can also download the output artifact of a step queried from a workflow (to current directory for default)

step = wf.query_step(name="hello")
download_artifact(step.outputs.artifacts["bar"])

Modify dflow.s3_config to configure artifact repository settings globally.

Note: dflow retains the relative path of the uploaded file/directory with respect to the current directory during uploading. If file/directory outside current directory is uploaded, its absolute path is used as the relative path in the artifact. If you want a different directory structure in the artifact with the local one, you can make soft links and then upload.

3.1.3. Steps

Steps is another kind of OP template which is defined by its constituent steps instead of a container. It can be seen as a sub-workflow or a super OP template consisting of some smaller OPs. A steps includes an array of arrays of steps, e.g. [[s00,s01],[s10,s11,s12]], where inner array represent concurrent steps while outer array is sequential. One can declare input/output parameters/artifacts for a steps by

steps.inputs.parameters["msg"] = InputParameter()
steps.inputs.artifacts["foo"] = InputArtifact()
steps.outputs.parameters["msg"] = OutputParameter()
steps.outputs.parameters["bar"] = OutputArtifact()

Add a step to a steps just like for a workflow

steps.add(step)

steps can be used as the template to instantiate a bigger step just like script OP templates. Thus one can construct complex workflows of nested structure. One is also allowed to recursively use a steps as the template of a building bloack inside it self to achieve dynamic loop.

The output parameter of a steps can be set to come from a step of it by

steps.outputs.parameters["msg"].value_from_parameter = step.outputs.parameters["msg"]

Here, step must be contained in steps. For assigning output artifact for a steps, use

steps.outputs.artifacts["foo"]._from = step.outputs.parameters["foo"]

3.1.4. DAG

DAG is another kind of OP template which is defined by its constituent tasks and their dependencies. The usage of DAG is similar to that of Steps. To add a task to a dag, use

dag.add(task)

The usage of task is also similar to that of step. Dflow will automatically detect dependencies among tasks of a dag (from input/output relations). Additional dependencies can be declared by

task_3 = Task(..., dependencies=[task_1, task_2])

3.1.5. Conditional step, parameters and artifacts

Set a step to be conditional by Step(..., when=expr) where expr is an boolean expression in string format. Such as "%s < %s" % (par1, par2). The step will be performed if the expression is evalutated to be true, otherwise skipped. The when argument is often used as the breaking condition of recursive steps. The output parameter of a steps (similar to dag) can be assigned as conditional by

steps.outputs.parameters["msg"].value_from_expression = if_expression(
    _if=par1 < par2,
    _then=par3,
    _else=par4
)

Similarly, the output artifact of a steps can be assigned as conditional by

steps.outputs.artifacts["foo"].from_expression = if_expression(
    _if=par1 < par2,
    _then=art1,
    _else=art2
)

3.1.6. Produce parallel steps using loop

In scientific computing, it is often required to produce a list of parallel steps which share a common OP template, and only differ in the input parameters. with_param and with_sequence are 2 arguments of Step for automatically generating a list of parallel steps. These steps share a common OP template, and only differ in the input parameters.

A step using with_param option generates parallel steps on a list (either a constant list or referring to another parameter, e.g. an output parameter of another step or an input parameter of the steps or DAG context), the parallelism equals to the length of the list. Each parallel step picks an item from the list by "{{item}}", such as

step = Step(
    ...
    parameters={"msg": "{{item}}"},
    with_param=steps.inputs.parameters["msg_list"]
)

A step using with_sequence option generates parallel steps on a numeric sequence. with_sequence is usually used in coordination with argo_sequence which returns an Argo's sequence. For argo_sequence, the number at which to start the sequence is specified by start (default: 0). One can either specify the number of elements in the sequence by count or the number at which to end the sequence by end. The printf format string can be specified by format to format the value in the sequence. Each argument can be passed with a parameter, argo_len which returns the length of a list may be useful. Each parallel step picks an element from the sequence by "{{item}}", such as

step = Step(
    ...
    parameters={"i": "{{item}}"},
    with_sequence=argo_sequence(argo_len(steps.inputs.parameters["msg_list"]))
)

3.1.7. Timeout

Set the timeout of a step by Step(..., timeout=t). The unit is second.

3.1.8. Continue on failed

Set the workflow to continue when a step fails by Step(..., continue_on_failed=True).

3.1.9. Continue on success number/ratio of parallel steps

For a group of parallel steps generated by with_param or with_sequence, set the workflow to continue when certain number/ratio of parallel steps succeed by Step(..., continue_on_num_success=n) or Step(..., continue_on_success_ratio=r).

3.1.10. Optional input artifacts

Set an input artifact to be optional by op_template.inputs.artifacts["foo"].optional = True.

3.1.11. Default value for output parameters

Set default value for an output parameter by op_template.outputs.parameters["msg"].default = default_value. The default value will be used when the expression in value_from_expression fails or the step is skipped.

3.1.12. Key of a step

One can assign a key for a step by Step(..., key="some-key") for the convenience of locating the step. The key can be regarded as an input parameter which may contain reference of other parameters. For instance, the key of a step can change with iterations of a dynamic loop. Once key is assigned to a step, the step can be query by wf.query_step(key="some-key"). If the key is unique within the workflow, the query_step method returns a list consist of only one element.

3.1.13. Resubmit a workflow

Workflows often have some steps that are expensive to compute. The outputs of previously run steps can be reused for submitting a new workflow. E.g. a failed workflow can be restarted from a certain point after some modification of the workflow template or even outputs of completed steps. For example, submit a workflow with reused steps by wf.submit(reuse_step=[step0, step1]). Here, step0 and step1 are previously run steps returned by query_step method. Before the new workflow runs a step, it will detect if there exists a reused step whose key matches that of the step about to run. If hit, the workflow will skip the step and set its outputs as those of the reused step. To modify outputs of a step before reusing, use step0.modify_output_parameter(par_name, value) for parameters and step0.modify_output_artifact(art_name, artifact) for artifacts.

3.1.14. Executor

For a "script step" (a step whose template is a script OP template), by default the Shell script or Python script runs in the container directly. Alternatively, one can modify the executor to run the script. Dflow offers an extension point for "script step" Step(..., executor=my_executor). Here, my_executor should be an instance of class derived from the abstract class Executor. An implementation class of Executor should implement a method render which converts original template to a new template.

class Executor(ABC):
    @abc.abstractmethod
    def render(self, template):
        pass

A context is similar to an executor, but assigned to a workflow Workflow(context=...) and affect every step.

3.1.15. Submit HPC/Bohrium job via dispatcher plugin

DPDispatcher is a python package used to generate HPC scheduler systems (Slurm/PBS/LSF) or Bohrium jobs input scripts and submit these scripts and poke until they finish. Dflow provides simple interface to invoke dispatcher as executor to complete script steps. E.g.

from dflow.plugins.dispatcher import DispatcherExecutor
Step(
    ...,
    executor=DispatcherExecutor(host="1.2.3.4",
        username="myuser",
        queue_name="V100")
)

For SSH authentication, one can either specify path of private key file locally, or upload authorized private key to each node (or equivalently add each node to the authorized host list). For configuring extra machine, resources or task parameters for dispatcher, use DispatcherExecutor(..., machine_dict=m, resources_dict=r, task_dict=t).

3.1.16. Submit Slurm job via virtual node

Following the installation steps in the wlm-operator project to add Slurm partitions as virtual nodes to Kubernetes (use manifests configurator.yaml, operator-rbac.yaml, operator.yaml in this project which modified some RBAC configurations)

$ kubectl get nodes
NAME                            STATUS   ROLES                  AGE    VERSION
minikube                        Ready    control-plane,master   49d    v1.22.3
slurm-minikube-cpu              Ready    agent                  131m   v1.13.1-vk-N/A
slurm-minikube-dplc-ai-v100x8   Ready    agent                  131m   v1.13.1-vk-N/A
slurm-minikube-v100             Ready    agent                  131m   v1.13.1-vk-N/A

Then you can assign a step to be executed on a virtual node (i.e. submit a Slurm job to the corresponding partition to complete the step)

step = Step(
    ...
    executor=SlurmJobTemplate(
        header="#!/bin/sh\n#SBATCH --nodes=1",
        node_selector={"kubernetes.io/hostname": "slurm-minikube-v100"}
    )
)

3.1.17. Use resources in Kubernetes

A step can also be completed by a Kubernetes resource (e.g. Job or custom resources). At the beginning, a manifest is applied to Kubernetes. Then the status of the resource is monitered until the success condition or the failure condition is satisfied.

class Resource(ABC):
    action = None
    success_condition = None
    failure_condition = None
    @abc.abstractmethod
        pass

3.1.18. Important note: variable names

Dflow has following restrictions on variable names.

Variable name Static/Dynamic Restrictions Example
Workflow/OP template name Static Lowercase RFC 1123 subdomain (must consist of lower case alphanumeric characters, '-' or '.', and must start and end with an alphanumeric character my-name
Step/Task name Static Must consist of alpha-numeric characters or '-', and must start with an alpha-numeric character My-name1-2, 123-NAME
Parameter/Artifact name Static Must consist of alpha-numeric characters, '_' or '-' my_param_1, MY-PARAM-1
Key name Dynamic Lowercase RFC 1123 subdomain (must consist of lower case alphanumeric characters, '-' or '.', and must start and end with an alphanumeric character my-name

3.1.19. Debug mode: dflow independent of Kubernetes

The debug mode is enabled by setting

from dflow import config
config["mode"] = "debug"

Before running a workflow locally, make sure that the dependencies of all OPs in the workflow are well-configured in the locally environment, unless the dispatcher executor is employed to submit jobs to some remote environments. The debug mode uses the current directory as the working directory by default. Each workflow will create a new directory there, whose structure will be like

python-lsev6
├── status
└── step-penf5
    ├── inputs
    │   ├── artifacts
    │   │   ├── dflow_python_packages
    │   │   ├── foo
    │   │   └── idir
    │   └── parameters
    │       ├── msg
    │       └── num
    ├── log.txt
    ├── outputs
    │   ├── artifacts
    │   │   ├── bar
    │   │   └── odir
    │   └── parameters
    │       └── msg
    ├── phase
    ├── script
    ├── type
    └── workdir
        ├── ...

The top level contains the status and all steps of the workflow. The directory name for each step will be its key if provided, or generated from its name otherwise. The step directory contains the input/output parameters/artifacts, the type and the phase of the step. For a step of type "Pod", its directory also includes the script, the log file and the working directory for the step.

3.1.20. Artifact storage plugins

The default storage for artifacts in dflow is a Minio deployment in the Kubernetes cluster. While other artifact storages are supported (e.g. Aliyun OSS, Azure blob storage (ABS), Google cloud storage(GCS)). Dflow provides an extension point to use customized storage in the artifact management. A storage client is a class implementing 5 abstract methods, upload, download, list, copy and get_md5 (optional), which offer the functionality of uploading file, downloading file, listing files with a particular prefix, copying file on the server side and getting the MD5 sum of file, respectively. Use a custom storage client object by configuring s3_config["storage_client"].

class StorageClient(ABC):
    @abc.abstractmethod
    def upload(self, key: str, path: str) -> None:
        pass
    @abc.abstractmethod
    def download(self, key: str, path: str) -> None:
        pass
    @abc.abstractmethod
    def list(self, prefix: str, recursive: bool = False) -> List[str]:
        pass
    @abc.abstractmethod
    def copy(self, src: str, dst: str) -> None:
        pass
    @abc.abstractmethod
    def get_md5(self, key: str) -> str:
        pass

3.2. Interface layer

3.2.1. Slices

In coordination with parallel steps, Slices helps user to slice input parameters/artifacts (which must be lists) to feed parallel steps and stack their output parameters/artifacts to lists in the same pattern. The Python OP only need to handle one slice. For example,

step = Step(name="parallel-tasks",
    template=PythonOPTemplate(
        ...,
        slices=Slices("{{item}}",
            input_parameter=["msg"],
            input_artifact=["data"],
            output_artifact=["log"])
    ),
    parameters = {
        "msg": msg_list
    },
    artifacts={
        "data": data_list
    },
    with_param=argo_range(5)
)

In this example, each item in msg_list is passed to a parallel step as the input parameter msg, each part in data_list is passed to a parallel step as the input artifact data. Finally, the output artifacts log of all parallel steps are collected to one artifact step.outputs.artifacts["log"].

It should be noticed that this feature by default passes full input artifacts to each parallel step which may only use some slices of these artifacts. In comparison, the subpath mode of slices only passes one single slice of the input artifacts to each parallel step. To use the subpath mode of slices,

step = Step(name="parallel-tasks",
    template=PythonOPTemplate(
        ...,
        slices=Slices(sub_path=True,
            input_parameter=["msg"],
            input_artifact=["data"],
            output_artifact=["log"])
    ),
    parameters = {
        "msg": msg_list
    },
    artifacts={
        "data": data_list
    })

Here, the slice pattern ({{item}}) of PythonOPTemplate and the with_param argument of the Step need not to be set, because they are fixed in this mode. Each input parameter and artifact to be sliced must be of the same length, and the parallelism equals to this length. Another noticeable point is that in order to use the subpath of the artifacts, these artifacts must be saved without compression when they are generated. E.g. declare Artifact(..., archive=None) in the output signs of Python OP, or specify upload_artifact(..., archive=None) while uploading artifacts. Besides, one can use dflow.config["archive_mode"] = None to set default archive mode to no compression globally.

3.2.2. Retry and error handling

Dflow catches TransientError and FatalError thrown from OP. User can set maximum number of retries on TransientError by PythonOPTemplate(..., retry_on_transient_error=n). Timeout error is regarded as fatal error for default. To treat timeout error as transient error, set PythonOPTemplate(..., timeout_as_transient_error=True). When a fatal error is raised or the retries on transient error reaches maximum retries, the step is considered as failed.

3.2.3. Progress

A OP can update progress in the runtime so that user can track its real-time progress

class Progress(OP):
    progress_total = 100
    ...
    def execute(op_in):
        for i in range(10):
            self.progress_current = 10 * (i + 1)
            ...

3.2.4. Upload python packages for development

To avoid frequently making image during development, dflow offers an interface to upload local packages into container and add them to $PYTHONPATH, such as PythonOPTemplate(..., python_packages=["/opt/anaconda3/lib/python3.9/site-packages/numpy"]). One can also globally specify packages to be uploaded, which will affect all OPs

from dflow.python import upload_packages
upload_packages.append("/opt/anaconda3/lib/python3.9/site-packages/numpy")

Releases

No releases published

Packages

No packages published

Languages

  • Python 90.4%
  • Jupyter Notebook 6.8%
  • Shell 2.8%