Timeline for How to ensure data consistency in system with multiple databases?
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 22, 2023 at 11:16 | comment | added | JonasH | @IanGoldby Then we are in agreement. I did not mean to imply that microservices could solve consistency problems, Only that a well designed system could reduce the chance, and mitigate the effects of such problems. I hope the text makes this clear, and apologize for any confusion. | |
| Sep 22, 2023 at 10:05 | comment | added | Ian Goldby | I don't think monolithic systems are more fault tolerant. I was only pointing out that switching to microservices is unlikely to make consistency problems go away. I think your edits make that clear too. | |
| Sep 22, 2023 at 9:41 | comment | added | Ian Goldby | Dealing with message non-delivery (and even failure of a listener while it is processing a message) is easy enough. But guarding against out-of-band failure modes (such as a botched upgrade/migration of the MQ infrastructure) is almost impossible. So you still need fault tolerance. | |
| Sep 22, 2023 at 9:15 | history | edited | JonasH | CC BY-SA 4.0 | added 742 characters in body |
| Sep 22, 2023 at 7:19 | comment | added | Ian Goldby | The microservice architecture is good for maintainability and scalability, but it's terrible for consistency - how does microservice A know when microservice B deletes a row that A has a reference to? So B broadcasts a delete event, but messages occasionally don't get delivered. The only answer is to design systems that are fault-tolerant. | |
| Sep 20, 2023 at 15:12 | history | answered | JonasH | CC BY-SA 4.0 |