71 lines
3.8 KiB
Markdown
71 lines
3.8 KiB
Markdown
---
|
||
layout: default
|
||
title: Managing ML models in OpenSearch Dashboards
|
||
nav_order: 120
|
||
redirect_from:
|
||
- /ml-commons-plugin/ml-dashbaord/
|
||
---
|
||
|
||
# Managing ML models in OpenSearch Dashboards
|
||
**Generally available 2.9**
|
||
{: .label .label-purple }
|
||
|
||
Administrators of machine learning (ML) clusters can use OpenSearch Dashboards to manage and check the status of ML models running inside a cluster. This can help ML developers provision nodes to ensure their models run efficiently.
|
||
|
||
You can register and deploy models using the API only. For more information, see [Using ML models within OpenSearch]({{site.url}}{{site.baseurl}}/ml-commons-plugin/model-serving-framework/).
|
||
|
||
## Enabling ML in Dashboards
|
||
|
||
In OpenSearch 2.6, ML functionality is disabled by default. To enable it, you need to edit the configuration in `opensearch_dashboards.yml` and then restart your cluster.
|
||
|
||
To enable the feature:
|
||
|
||
1. In your OpenSearch cluster, navigate to your Dashboards home directory; for example, in Docker, `/usr/share/opensearch-dashboards`.
|
||
2. Open your local copy of the Dashboards configuration file `opensearch_dashboards.yml`. If you don't have a copy, get one from GitHub: [`opensearch_dashboards.yml`](https://github.com/opensearch-project/OpenSearch-Dashboards/blob/main/config/opensearch_dashboards.yml).
|
||
3. Add the setting `ml_commons_dashboards.enabled:` to `opensearch_dashboards.yml`. Then, set it to `ml_commons_dashboards.enabled: true` and save the configuration file.
|
||
4. Restart the Dashboards container.
|
||
5. Verify that the feature configuration settings were created and configured properly by launching OpenSearch Dashboards. The Machine Learning section should appear under **OpenSearch plugins**.
|
||
|
||
## Accessing ML functionality in Dashboards
|
||
|
||
To access ML functionality in OpenSearch Dashboards,select **OpenSearch plugins** > **Machine Learning**.
|
||
|
||
<img src="{{site.url}}{{site.baseurl}}/images/ml/ml-dashboard/ml-dashboard.png" alt="Machine Learning section in OpenSearch dashboards">
|
||
|
||
In the Machine Learning section, you now have access to the **Deployed models** dashboard.
|
||
|
||
## Deployed models dashboard
|
||
|
||
The deployed models dashboard gives admins the ability to check the status of any models stored inside your OpenSearch cluster.
|
||
|
||
<img src="{{site.url}}{{site.baseurl}}/images/ml/ml-dashboard/deployed-models.png" alt="The deployed models view.">
|
||
|
||
The dashboard includes the following information about the model:
|
||
|
||
- **Name**: The name of the model given upon upload.
|
||
- **Status**: The number of nodes for which the model responds.
|
||
- When all nodes are responsive, the status is **Green**.
|
||
- When some nodes are responsive,the status is **Yellow**.
|
||
- When all nodes are unresponsive,the status is **Red**.
|
||
- **Model ID**: The model ID.
|
||
- **Action**: What actions you can take with the model.
|
||
|
||
As of OpenSearch 2.6, the only action available is **View Status Details**, shown in the following image.
|
||
|
||
<img src="{{site.url}}{{site.baseurl}}/images/ml/ml-dashboard/view-status-details.png" alt="You can view status details under actions.">
|
||
|
||
When selected, the Status Details panel appears.
|
||
|
||
The panel provides the following details inside the panel:
|
||
|
||
- **Model ID**
|
||
- **Model status by node**: The number of nodes for which the model is responsive.
|
||
|
||
A list of nodes gives you a view of each node the model is running on, including each node’s **Node ID** and status, as shown in the following image. This is useful if you want to use the node's **Node ID** to determine why a node is unresponsive.
|
||
|
||
<img src="{{site.url}}{{site.baseurl}}/images/ml/ml-dashboard/model-node-details.png" alt="The status of each node running the model.">
|
||
|
||
## Next steps
|
||
|
||
For more information about how to manage ML models in OpenSearch, see [Using ML models within OpenSearch]({{site.url}}{{site.baseurl}}/ml-commons-plugin/model-serving-framework/).
|