How do you handle the timeout in independent microservices?
I have a request that passes over 3 microservices before reaching the destination, but the time en each microservice (approx 800ms) is reflecting a bad UX. Can my architecture be improved?
does it go trough 3 or just have to reach 3? because this can be done by a queue.
Besides that maybe a promise / queue system could help in general since you can say "we get back to you" but there is instant feedback? But this already is very speculative :) since I would have to analyze it to say anything besides classic approaches "it's slow make it async" it's slow "use a parallel system instead of a serial one" :)
The system is for a chat... the first microservice catch the message and authenticate if the user sender is a sender autenticated, the second message make sentiment analisys and 3rd service fetch a database based on sentiment analysis forn back microservice, and build apropiate response for a client in the chat... That is the why I not see queue in this apporach, I know that all can be toguether in a single service, but in next month we think add more funcionality to this chatbot, for example ours customers wants to connect to a Google Sheets and not to a database (I know that sheets is not a database is only an example.)... thanks...
j
stuff ;)
does it go trough 3 or just have to reach 3? because this can be done by a queue.
Besides that maybe a promise / queue system could help in general since you can say "we get back to you" but there is instant feedback? But this already is very speculative :) since I would have to analyze it to say anything besides classic approaches "it's slow make it async" it's slow "use a parallel system instead of a serial one" :)