Add a new query parameter, similar to maxTimeMS, with the purpose of limiting how long an operation runs from driver/client point of view.

It could be a different parameter which acts like maxTimeMS on the backend and which acts like socket timeout on the client side.

Socket timeout is a global setting and it does not work well when different operations have different SLA's within the application. maxTimeMS also is not reliable from client point of view as it can take longer than maxTimeMS to get an error.

Here is the discussion related to this with Server team : https://jira.mongodb.org/browse/SERVER-31578