I don't understand this article. Which best practices are you talking about? Why would you make an invalid message visible to other receivers after a lock timeout? Can you explain a little better?
All good questions, thanks! I understand that this post might be difficult to digest. The first part is a walk through on a use case inspired from real life where an invalid message was published to a processing queue (why? due to a bug in the developer's code) and caused continuous looping between the queue and the listener of that queue.
The best practices I highlighted in a subtle manner are:
Cool, that does help my understanding. It's worth keeping in mind that you can't rely on message validation before it is published to the queue (think malicious user!), so having the receiver perform validation to pass any invalid messages to a dead letter queue is essential.
I don't understand this article. Which best practices are you talking about? Why would you make an invalid message visible to other receivers after a lock timeout? Can you explain a little better?
All good questions, thanks! I understand that this post might be difficult to digest. The first part is a walk through on a use case inspired from real life where an invalid message was published to a processing queue (why? due to a bug in the developer's code) and caused continuous looping between the queue and the listener of that queue.
The best practices I highlighted in a subtle manner are:
Hope this clarifies it.
Cool, that does help my understanding. It's worth keeping in mind that you can't rely on message validation before it is published to the queue (think malicious user!), so having the receiver perform validation to pass any invalid messages to a dead letter queue is essential.
Keep up the writing!
Excellent, glad to hear! Thanks, will do! :)