search

MicrosoftDocs / azure-docs

Open source documentation of Microsoft Azure
https://docs.microsoft.com/azure
Creative Commons Attribution 4.0 International
10.26k stars 21.43k forks source link

Cache configuration for API 3.x #103825

Closed yutaro113 closed 1 year ago

yutaro113 commented 1 year ago

Hi Team,

Here it's mentioned that external cache needs to be configured. https://learn.microsoft.com/en-us/azure/cognitive-services/computer-vision/deploy-computer-vision-on-premises#configure-helm-chart-values-for-deployment image

But in the below doc section which is linked from above document, "Cache:Redis" option seems valid only for API v2.x. how should it be configured for API v3.x? This point is not mentioned so feel confusing. should it be mentioned in those docs? https://learn.microsoft.com/en-us/azure/cognitive-services/computer-vision/computer-vision-resource-container-config?tabs=version-3-2#configuration-settings

Kindly confirm. Regards, Yutaro

Document Details

⚠ Do not edit this section. It is required for learn.microsoft.com ➟ GitHub issue linking.

YashikaTyagii commented 1 year ago

@yutaro113 Thanks for your feedback! We will investigate and update as appropriate.

Naveenommi-MSFT commented 1 year ago

@yutaro113 Thank you for bringing this to our attention. I've delegated this to content author @aahill, who will review it and offer his insightful opinions.

aahill commented 1 year ago

Reassigning to @PatrickFarley who works on Computer vision

reassign:PatrickFarley

leon-vd commented 1 year ago

@yutaro113 From what I can tell, you need to use an Azure Storage Account (Blob and Queue) for this in version 3.x. The docs don't mention what format the values for these parameters should be, so I figured I'd share how I got it working. The bit that tripped me up for a while was not including 'QueueEndpoint=' in the value.

I only got it working with using SAS tokens when testing this out. I'm unsure if you can do it using one of the other storage account authentication methods. If you get that working, feel free to share.

The below worked for me. It creates a whole bunch of queues in the storage account and also a container called 'read', where it stores all the results under. I was able to do an async request on one container and retrieve it on another container running on a different port.

docker run -p 5000:5000 \ mcr.microsoft.com/azure-cognitive-services/vision/read:3.2-model-2022-04-30 \ Eula=accept \ Billing={ENDPOINT_URI} \ ApiKey={API_KEY} \ Queue:Azure:ConnectionString="QueueEndpoint=https://.queue.core.windows.net/?" \ Storage:ObjectStore:AzureBlob:ConnectionString="BlobEndpoint=https://.blob.core.windows.net/?"

PatrickFarley commented 1 year ago

@yutaro113 Did that answer your question? I'm a bit out of my depth here

PatrickFarley commented 1 year ago

please-close