Solution Question - UX

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
Report Content as Inappropriate

Solution Question - UX

This post has NOT been accepted by the mailing list yet.
Hi all! New to Rabbit and new to the forum!

I've got an architecture discussion. We're building a web based process that I'd like some "peer review" from the Rabbit experts here. The process as it stands now is:
1) User hits a page and several systems are queried for data. These systems have different aspects (performance, processing time, etc)
2) Once the data is displayed the customer can change the data and submit it. At this point we're planning to publish this to Rabbit and have the underlying systems subscribe to durable queues. So far so good.
3) So now say the user immediately changes their mind and decides to tweak the changes. They go back in and now they see (query) the old state as the process is async and its possible that one of the systems has't completed the processing yet. Great - we're looking to solve this by having a Redis cache hold the latest changes and show that as the current state until the back end message have been processed.

So now how do we know to clear the cache and re-query the underlying system for the updated state? Do we have a process in our service layer in front of Rabbit that query's Rabbit for a correlationID to clear the old cached items? What technologies exist in Rabbit that could help support this? I hope this is clear. This seems like a typical pattern for Web+ Rabbit. Thanks!