I would like to have a better understanding of the side effects on using blocking calls from inside software timers callback functions. I guess one of the conssequences is a delay on executing the next timer event, since all events are processed on a single task, right? But, besides this delay, are there any other side effects that can disrupt code execution?
The biggest effect is that it can delay other timer/pended function events.
A second possible effect is that it blocks the timer task from processing its queue, so if too many tasks queue up new requests those will block, and possibly timeout.
If the timer task is blocked it will not be removing items from the
timer command queue, so the queue could get full.
Copyright (C) Amazon Web Services, Inc. or its affiliates. All rights reserved.