Deep dive in Durable Functions - Laurent Bugnion - NDC Oslo 2021

Azure Functions are one of the offerings implemented by Microsoft for serverless technologies. While they are extremely easy and fast to deploy and run, and offer a lot of versatility to implement backend functionalities, they also have some limitations. This is why Durable Functions were developed.
With Durable Functions, you have even more power to create powerful backend systems. You can have long running functions, massively parallel processing, and other application patterns that we will review and demo. But all this power comes with some caveats, and it is very important to understand precisely how Durable Functions work, and what the underlying mechanisms are. In this session, Laurent Bugnion, a Microsoft Cloud Advocate since 2017 will take you to a deep dive into Azure Durable Functions. You will get out of this talk with a deeper understanding of Azure Functions, the Durable extensions and some demos illustrating all the points.
Check out more of our featured speakers and talks at
ndcconferences.com/
ndcoslo.com/

[@aprobinda]

42:10 that is exactly the problem I had a few days ago! I learned about serialization the hard way: a class parameter in an activity function must have public setters in all properties.
Thank you for this video!

[@tomvandenberg4210]

Very good deep dive, the point at 50:00 gave my use case more convidence

[@logaraj6938]

It was awesome session for Azure paas developer

[@glob3rek]

Thanks for the overview, nice and slick

[@signupwithful]

One of the best ADF videos

[@cyndijohnson5473]

That acronym is overloaded: aka Azure Data Factory

[@SFORSANDEEP]

Wawwww.. Super class..

[@bfg5244]

you could also have http-triggered (admin) azure function that executes orchestrationClient.PurgeInstanceHistoryAsync to ease developing

[@bfg5244]

not to mention you could totally debug Durable functions locally with Azurite

[@paulkersey2424]

English please

[@9888622400]

Hi, can you share any approach in python as well?