Skip to content

Commit f41135c

Browse files
committed
Add mirrored traffic for studio
1 parent 3845a10 commit f41135c

File tree

5 files changed

+23
-3
lines changed

5 files changed

+23
-3
lines changed

articles/machine-learning/how-to-safely-rollout-online-endpoints.md

Lines changed: 23 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -633,8 +633,27 @@ After testing, you can set the mirror traffic to zero to disable mirroring:
633633

634634
# [Studio](#tab/azure-studio)
635635

636-
The studio doesn't support mirrored traffic. See the Azure CLI or Python tabs for steps to mirror traffic to a deployment.
636+
To mirror 10% of the traffic to the `green` deployment:
637637

638+
1. From the endpoint Details page, Select **Update traffic**.
639+
1. Slide the button to **Enable mirrored traffic (Preview)**.
640+
1. Select the **green** deployment in the "Deployment name" dropdown menu.
641+
1. Keep the default traffic allocation of 10%.
642+
1. Select **Update**.
643+
644+
:::image type="content" source="media/how-to-safely-rollout-managed-endpoints/mirror-traffic-to-green-deployment.png" alt-text="Screenshot showing how to mirror a percentage of traffic to the green deployment." lightbox="media/how-to-safely-rollout-managed-endpoints/mirror-traffic-to-green-deployment.png":::
645+
646+
The endpoint details page now shows mirrored traffic allocation of 10% to the `green` deployment.
647+
648+
:::image type="content" source="media/how-to-safely-rollout-managed-endpoints/endpoint-details-showing-mirrored-traffic-allocation.png" alt-text="Endpoint details page showing mirrored traffic allocation in the deployment summary." lightbox="media/how-to-safely-rollout-managed-endpoints/endpoint-details-showing-mirrored-traffic-allocation.png":::
649+
650+
<!-- M.A.: Test mirrored traffic and confirm that the specific percentage of traffic was mirrored to the green deployment -->
651+
652+
After testing, you can disable mirroring:
653+
654+
1. From the endpoint Details page, Select **Update traffic**.
655+
1. Slide the button next to **Enable mirrored traffic (Preview)** again to disable mirrored traffic.
656+
1. Select **Update**.
638657
---
639658

640659
## Allocate a small percentage of live traffic to the new deployment
@@ -707,7 +726,7 @@ Use the following steps to delete an individual deployment from a managed online
707726
# [Studio](#tab/azure-studio)
708727

709728
> [!NOTE]
710-
> You cannot delete a deployment that has traffic allocated to it. You must first [set traffic allocation](#send-all-traffic-to-your-new-deployment) for the deployment to 0% before deleting it.
729+
> You cannot delete a deployment that has live traffic allocated to it. You must first [set traffic allocation](#send-all-traffic-to-your-new-deployment) for the deployment to 0% before deleting it.
711730
712731
1. In the endpoint Details page, find the blue deployment.
713732
1. Select the **delete icon** next to the deployment name.
@@ -737,13 +756,14 @@ If you aren't going to use the endpoint and deployment, you should delete them.
737756
1. Select an endpoint by checking the circle next to the model name.
738757
1. Select **Delete**.
739758

740-
Alternatively, you can delete a managed online endpoint directly by selecting the **Delete** icon in the [endpoint details page](#view-managed-online-endpoints).
759+
Alternatively, you can delete a managed online endpoint directly by selecting the **Delete** icon in the endpoint's details page.
741760

742761
---
743762

744763
## Next steps
745764
- [Explore online endpoint samples](https://github.com/Azure/azureml-examples/tree/v2samplesreorg/sdk/python/endpoints)
746765
- [Deploy models with REST](how-to-deploy-with-rest.md)
766+
- [Use network isolation with managed online endpoints](how-to-secure-online-endpoint.md)
747767
- [Access Azure resources with a online endpoint and managed identity](how-to-access-resources-from-endpoints-managed-identities.md)
748768
- [Monitor managed online endpoints](how-to-monitor-online-endpoints.md)
749769
- [Manage and increase quotas for resources with Azure Machine Learning](how-to-manage-quotas.md#azure-machine-learning-managed-online-endpoints)
Loading
Loading
116 KB
Loading
2.68 KB
Loading

0 commit comments

Comments
 (0)