The Brand New Way to do Fault Handling in Polly V8

[@MilanJovanovicTech]

Want to master Clean Architecture? Go here: bit.ly/3PupkOJ Want to unlock Modular Monoliths? Go here: bit.ly/3SXlzSt

[@pepega8572]

Bro, your content is amazing. This is the second time I've needed to figure out how to use new library, and you help me with this. Thanks!) Keep doing this way)

[@MilanJovanovicTech]

Glad it was helpful! :)

[@nepalxplorer]

The Performance Improvement is literally impressive (mainly 0 allocation)

[@MilanJovanovicTech]

They did a great job 😁

[@מתןשולמן]

great video, Milan. Your explanations are always top-notch. Thank you and keep up the good work!

[@MilanJovanovicTech]

Glad you like them!

[@kodindoyannick5328]

Amazing content! Thanks Milan for your sharing.

[@MilanJovanovicTech]

My pleasure!

[@richardaubin1951]

I've been playing around with this new version for a few hours now, and it's much easier to grasp then the previous version. Could you have a go at a video on the Dynamic Reload functionality? I think it would be good to see that in action for concurrent rate limit, retry and circuit breaker options.

[@MilanJovanovicTech]

Will take a look at it!

[@tjagusz]

@@MilanJovanovicTech I just found this video, a circuit breaker would be awesome, especially in a microservice architecture, when for example you have a lot of messages to process and you are relying on an external API that can fail, so instead of only retrying we should stop trying for a specific time. Looking forward to this content in the next video ;) BTW another useful video. Thank you!

[@Sammy2100]

I am switching to this ASAP. Thanks for sharing.

[@MilanJovanovicTech]

Check out this: www.milanjovanovic.tech/blog/building-resilient-cloud-applications-with-dotnet

[@Sammy2100]

@@MilanJovanovicTech yes, I got this email notification and found this video from it. i have updated my .net8 code using this and its awesome. Thanks.

[@antonmartyniuk]

Interesting new concept of the library. Gonna look how the new http extensions package looks like, in most cases I prefer adding policies (strategies) when registering http client as factory in DI. The important thing to mention: Polly v8 doesn't need a Wrap to combine policies, it combines them automatically. And othe order is FIFO instead of FILO in Wrap policy in v7

[@MilanJovanovicTech]

I'll have to check for wrapping. The one test I did I got some strange behavior in the end.

[@antonmartyniuk]

@@MilanJovanovicTech I read about wrapping on the official site

[@okcharles7]

As I was introduced to this kind of technique for the first time by this video, I just wonder what is the performance benefit over try-catch.

[@MilanJovanovicTech]

The benefit is fault handling and recovering from the exception - not necessarily the performance

[@yasirirfan5262]

excellent work.

[@MilanJovanovicTech]

Many thanks!

[@vikas4483]

Could you please make a video on best practices with sample examples on exception handling in mvc core applications

[@MilanJovanovicTech]

Check this out for the time being: kzbin.info/www/bejne/fmSok5mipb2Jo9E

[@vikas4483]

@@MilanJovanovicTech Thanks 🙏

[@PelFox]

Can you still hook it into service container for HttpClient? That's how I've generally used it to handle 5xx errors for all added clients.

[@MilanJovanovicTech]

Yes, you can still use the Polly extensions package

[@baranacikgoz]

Impressive video.

[@MilanJovanovicTech]

Thank you very much!

[@tiagosantos2136]

In a clean architecture, where is the best place to put resilience strategies? in endpoints, applicatio layer, infra?

[@MilanJovanovicTech]

Probably infra

[@josefromspace]

Great video, thank you very much.

[@MilanJovanovicTech]

You are welcome!

[@PKMuhammedSaeed]

Please make a video on latest Microsoft.Extensions.Resilience library

[@MilanJovanovicTech]

Good suggestion

[@claudiovalerio7253]

Is it just me or the new builder stuff seems a step back? I mean, it probably would benefit some complex policy composition, but for simpler policies seems to be an awful lot of additional code needed to have the same result. Performance improvements are impressive, though...

[@MilanJovanovicTech]

I don't mind it, since you probably define these policies in one place and reuse them

[@squadwuschel]

How about unit Testing an Mocking the new Poly plicies in your code?

[@lettuceturnipthebeets790]

since it supports DI and is fully based on instances now, I think it makes it even easier to prep and test the policies honestly :0

[@MilanJovanovicTech]

Isn't it easier than with a static class?

[@W1ese1]

@@MilanJovanovicTech Question is: should you even mock the policies? I'd say no. Because in reality you want to test the scenario where the code fails for which you use the polly pipelines

[@MsbowElite]

Please create a video building Unit test for http clients and retry policies.

[@MilanJovanovicTech]

Good idea

[@antondoit]

Great

[@MilanJovanovicTech]

Thank you!

[@exeC13]

Using magic strings is always something that worries me. "How many refs does your di-pipeline-provider have / is it used at all?". [Ctrl]+[Shift]+[F] to the rescue 😖 "Great, no compile errors" ... Runtime steps in and alerts you about a typo in your magic string 🎉

[@MilanJovanovicTech]

It's a demo damn it... Of course you will use constants 😁

[@santoshr4212]

over confusing. Static version was lot easier

[@MilanJovanovicTech]

They did it for the other optimizations they get from the new approach

[@AlexDresko]

Easier != Better

[@aakashamatya3580]

The syntax has become too verbose.

[@MilanJovanovicTech]

Yes. We'll see where they take it.

[@ronaldschutte7948]

Way to complicated

[@MilanJovanovicTech]

Not much different. We'll see where they go from here.

[@techpc5453]

[@MilanJovanovicTech]

👋