importing from visual studio online, full benchmark #53
Conversation
|
|
||
| Follow this guide to get up and running with Python Durable Functions! | ||
|
|
||
| > **Note:** (2020-06-09) Because updated versions of the Azure Functions Python worker, extension bundles, and templates that are needed for the preview have not yet been published or deployed to Azure, you need to use a dev container with Visual Studio Code Remote for Containers or Visual Studio Online. |
There was a problem hiding this comment.
Why do we need these containers? Is it necessary to build and deploy the function or do you need to provide Azure with these containers to run the function in the cloud?
| tasks = [] | ||
| for file_name in frame_file_names: | ||
| tasks.append(context.call_activity("detectTemplate", file_name)) | ||
| out_list = yield context.task_all(tasks) |
There was a problem hiding this comment.
Do you understand it correctly: this calls the function splitVideo, then calls detectTemplate in parallel using the output of previous invocation as an input to the concurrenct executions?
| import json | ||
| from urllib.request import urlopen | ||
| # templatePath = "https://severless-test-aleqsio.s3.amazonaws.com/template.png" | ||
| templatePath ="https://storageaccountbaserbb3d.blob.core.windows.net/test01/template.png" |
There was a problem hiding this comment.
Similar to comments on review #47. Template should be part of benchmark resources. It either becomes the part of function resources (see this for an example - html template is packed with code), or is uploaded (example). For small files I'd recommend the former.
| { | ||
| "IsEncrypted": false, | ||
| "Values": { | ||
| "AzureWebJobsStorage": "DefaultEndpointsProtocol=https;AccountName=storageaccountbaserbb3d;AccountKey=+K2l9lf3TK0fUzacwM1wmAlbsgyar4KXLMJtZ1QN/yzIrwTdWM3dlG04QNlWLY7u62jyFgv/a8WR80jMv5mhIg==;EndpointSuffix=core.windows.net", |
There was a problem hiding this comment.
Is this your storage account hardcoded? If yes, then shouldn't this file be generated dynamically when storage account is resolved?
| @@ -0,0 +1,4 @@ | |||
| { | |||
| "$schema": "http://json.schemastore.org/proxies", | |||
|
|
||
| def current_milli_time(): return int(round(time.time() * 1000)) | ||
| # path = "https://severless-test-aleqsio.s3.amazonaws.com/video.mp4" | ||
| path = 'https://storageaccountbaserbb3d.blob.core.windows.net/test01/video.mp4' |
There was a problem hiding this comment.
As above - that's the input file? Then use input.py so it's automatically used as input.
|
Closing after a long period of inactivity - Azure Durable functions should be provided in the new PR #88 |
Azure benchmark similar to 001.video-template-matching on AWS using experimental python azure durable