Skip to content

Commit

Permalink
MemVerge MMCloud Agent (#1128)
Browse files Browse the repository at this point in the history
* Add MMCloud plugin to Integrations

Signed-off-by: Edwin Yu <[email protected]>

* Remove default image from README and example

Signed-off-by: Edwin Yu <[email protected]>

* Fix end of requirements.in

Signed-off-by: Edwin Yu <[email protected]>

* update requirement.txt for mmcloud_plugin

Signed-off-by: helenzhangyc <[email protected]>

---------

Signed-off-by: Edwin Yu <[email protected]>
Signed-off-by: helenzhangyc <[email protected]>
Co-authored-by: helenzhangyc <[email protected]>
  • Loading branch information
edwinyyyu and helenzhangyc authored Oct 2, 2023
1 parent 0ca1523 commit fe4968d
Show file tree
Hide file tree
Showing 8 changed files with 510 additions and 0 deletions.
1 change: 1 addition & 0 deletions docs/index.md
Original file line number Diff line number Diff line change
Expand Up @@ -448,6 +448,7 @@ auto_examples/greatexpectations_plugin/index
auto_examples/hive_plugin/index
auto_examples/k8s_pod_plugin/index
auto_examples/mlflow_plugin/index
auto_examples/mmcloud_plugin/index
auto_examples/modin_plugin/index
auto_examples/kfmpi_plugin/index
auto_examples/onnx_plugin/index
Expand Down
2 changes: 2 additions & 0 deletions docs/integrations.md
Original file line number Diff line number Diff line change
Expand Up @@ -104,6 +104,8 @@ the Flyte task that use the respective plugin.
- Running tasks and workflows on AWS batch service
* - {doc}`Hive <auto_examples/hive_plugin/index>`
- Run Hive jobs in your workflows.
* - {doc}`MMCloud <auto_examples/mmcloud_plugin/index>`
- Execute tasks using MemVerge Memory Machine Cloud
* - {doc}`Snowflake <auto_examples/snowflake_plugin/index>`
- Run Snowflake jobs in your workflows.
* - {doc}`Databricks <auto_examples/databricks_plugin/index>`
Expand Down
27 changes: 27 additions & 0 deletions examples/mmcloud_plugin/Dockerfile
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
FROM python:3.11-slim-bookworm
LABEL org.opencontainers.image.source https://github.com/flyteorg/flytesnacks

WORKDIR /root
ENV VENV /opt/venv
ENV LANG C.UTF-8
ENV LC_ALL C.UTF-8
ENV PYTHONPATH /root

WORKDIR /root

ENV VENV /opt/venv
# Virtual environment
RUN python3 -m venv ${VENV}
ENV PATH="${VENV}/bin:$PATH"

# Install Python dependencies
COPY requirements.txt /root
RUN pip install -r /root/requirements.txt

# Copy the actual code
COPY . /root

# This tag is supplied by the build script and will be used to determine the version
# when registering tasks, workflows, and launch plans
ARG tag
ENV FLYTE_INTERNAL_IMAGE $tag
80 changes: 80 additions & 0 deletions examples/mmcloud_plugin/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,80 @@
# Memory Machine Cloud

```{eval-rst}
.. tags:: AWS, GCP, AliCloud, Integration, Advanced
```

[MemVerge](https://memverge.com/) [Memory Machine Cloud](https://www.mmcloud.io/) (MMCloud)—available on AWS, GCP, and AliCloud—empowers users to continuously optimize cloud resources during runtime, safely execute stateful tasks on spot instances, and monitor resource usage in real time. These capabilities make it an excellent fit for long-running batch workloads.

Flyte can be integrated with MMCloud, allowing you to execute Flyte tasks using MMCloud.

## Installation

To install the plugin, run the following command:

```{eval-rst}
.. prompt:: bash
pip install flytekitplugins-mmcloud
```

To get started with MMCloud, refer to the [MMCloud User Guide](https://docs.memverge.com/mmce/current/userguide/olh/index.html).

## Configuring the backend to get MMCloud working

The MMCloud plugin is [enabled in FlytePropeller's configuration](https://docs.flyte.org/en/latest/deployment/plugins/memverge/mmcloud.html).

## Getting Started

This plugin allows executing `PythonFunctionTask` using MMCloud without changing any function code.

```{eval-rst}
.. testcode:: awsbatch-quickstart
from flytekitplugins.mmcloud import MMCloudConfig
@task(task_config=MMCloudConfig())
def to_str(i: int) -> str:
return str(i)
```

[Resource](https://docs.flyte.org/projects/cookbook/en/latest/auto_examples/productionizing/customizing_resources.html) (cpu and mem) requests and limits, [container](https://docs.flyte.org/projects/cookbook/en/latest/auto_examples/customizing_dependencies/multi_images.html) images, and [environment](https://docs.flyte.org/projects/flytekit/en/latest/generated/flytekit.task.html) variable specifications are supported.

[ImageSpec](https://docs.flyte.org/projects/cookbook/en/latest/auto_examples/customizing_dependencies/image_spec.html) may be used to define images to run tasks.

### Credentials

The following [secrets](https://docs.flyte.org/projects/cookbook/en/latest/auto_examples/productionizing/use_secrets.html) are required to be defined for the agent server:
* `mmc_address`: MMCloud OpCenter address
* `mmc_username`: MMCloud OpCenter username
* `mmc_password`: MMCloud OpCenter password

### Defaults

Compute resources:
* If only requests are specified, there are no limits.
* If only limits are specified, the requests are equal to the limits.
* If neither resource requests nor limits are specified, the default requests used for job submission are `cpu="1"` and `mem="1Gi"`, and there are no limits.

### Agent Image

Install `flytekitplugins-mmcloud` in the agent image.

A `float` binary (obtainable via the OpCenter) is required. Copy it to the agent image `PATH`.

Sample `Dockerfile` for building an agent image:
```dockerfile
FROM python:3.11-slim-bookworm

WORKDIR /root
ENV PYTHONPATH /root

# flytekit will autoload the agent if package is installed.
RUN pip install flytekitplugins-mmcloud
COPY float /usr/local/bin/float

CMD pyflyte serve --port 8000
```

```{auto-examples-toc}
example
```
Empty file.
42 changes: 42 additions & 0 deletions examples/mmcloud_plugin/mmcloud_plugin/example.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,42 @@
# %% [markdown]
# # Memory Machine Cloud
#
# This example shows how to use the MMCloud plugin to execute tasks on MemVerge Memory Machine Cloud.

# %%
from flytekit import Resources, task, workflow
from flytekitplugins.mmcloud import MMCloudConfig

# %% [markdown]
# `MMCloudConfig` configures `MMCloudTask`. Tasks specified with `MMCloudConfig` will be executed using MMCloud. Tasks will be executed with requests `cpu="1"` and `mem="1Gi"` by default.

# %%
@task(task_config=MMCloudConfig())
def to_str(i: int) -> str:
return str(i)


@task(task_config=MMCloudConfig())
def to_int(s: str) -> int:
return int(s)


# %% [markdown]
# [Resource](https://docs.flyte.org/projects/cookbook/en/latest/auto_examples/productionizing/customizing_resources.html) (cpu and mem) requests and limits, [container](https://docs.flyte.org/projects/cookbook/en/latest/auto_examples/customizing_dependencies/multi_images.html) images, and [environment](https://docs.flyte.org/projects/flytekit/en/latest/generated/flytekit.task.html) variable specifications are supported.

# %%
@task(
task_config=MMCloudConfig(submit_extra="--migratePolicy [enable=true]"),
requests=Resources(cpu="1", mem="1Gi"),
limits=Resources(cpu="2", mem="4Gi"),
environment={"KEY": "value"},
)
def concatenate_str(s1: str, s2: str) -> str:
return s1 + s2


@workflow
def concatenate_int_wf(i1: int, i2: int) -> int:
i1_str = to_str(i=i1)
i2_str = to_str(i=i2)
return to_int(s=concatenate_str(s1=i1_str, s2=i2_str))
2 changes: 2 additions & 0 deletions examples/mmcloud_plugin/requirements.in
Original file line number Diff line number Diff line change
@@ -0,0 +1,2 @@
flytekitplugins-envd
flytekitplugins-mmcloud
Loading

0 comments on commit fe4968d

Please sign in to comment.