Independant from http timeout

_{Photo by Louis Hansel on Unsplash}

We have a process to process a large amount of data, but it was bound to http timeout. And easily, we got 500 error.

This diagram shows original solution as a sequence diagram.

We decouple the data retrieval from http call. First we initiate the work, but send an id as response.
Then, we store the response as a file on Storage, and serve it when the frontend request it.

This diagram of the solution shows the sequence diagram on how to decouple generation of the file from sending it.

Hope this help !