Eldert Grootenboer presents 'Don't put your messages in a bottle; Implement messaging patterns' July 24th at Nebraska.Code().
https://nebraskacode.amegala.com/
#MessagingPatterns #modernsoftwarearchitecture #Nebraska #AzureServiceBus #Microsoft #dataconsistency #TechConference #softwaredevelopment #softwareengineering #technologyevents #TechTalk #networkingevent #sponsorshipopportunity #lincoln
Still no solution to this #azure #azureservicebus problem https://github.com/Azure/azure-sdk-for-net/issues/40773. As the SDK team refers to the Azure community feedback forum, here it is … https://feedback.azure.com/d365community/idea/0c0522b6-5a9e-ee11-a81c-000d3a05396f
@CodeAndCoffee Building out the feature wish-list for the game I'm thinking of building, then working on some POCs and subjecting them to load-testing to make sure my assumptions are correct about the tech stack and approach I want to use.
I'd like to create an event based, ledger type system.
My 1st thoughts are:
I need a pub/sub architecture but also a 'ledger' style, accounting-based system to track state.
Thoughts?
Updated to .NET 7 ASP.NET Core with Azure Service Bus
@nova I wrote a Charging Station Management System based on the Open Charging Point Protocol v1.6/2.0.1. It's written in #dotnet6 and hosted in #azure. I use #azureservicebus for communication between the different services. Frontend is a plain old #angular SPA. My wallbox connects via #websockets and sends and receives all kind of data to monitor and manage charging sessions. I mainly used it as a learning project to explore #dddesign a little further.
Do you need a #azure service bus explorer app for Mac? (The app is of course developed with .NET MAUI)
https://apps.apple.com/se/app/azservicebus/id6444410679?l=en&mt=12
Another way to say the above is this:
`Logs, with apppropriate notifications, are a perfectly valid Deadletter infrastructure for messages that cannot reasonably be reprocessed automatically.`
Deadletter queues or topics are only required in 2 cases:
1. When the message can be automatically processed in the future
2. When infrastructure constraints make them the simplest option
If a message cannot be automatically processed in the future (ie order needs to be maintained) it is better to log the failed message with appropriate notifications rather than putting it into deadletter infrastructure such as a queue.
