Add tests for Sentry logging

Icepeak logs errors to Sentry in case an exception occurs (see #52). It would be nice to see a number of tests that verify this functionality. These tests would somehow cause Icepeak to crash deliberately and verify that the correct errors are send to the Sentry DSN that is specified, for example by inspecting the records that are send from Icepeak. It should not be possible to let Icepeak crash when in production.

Start by reading issue #52 and locating Icepeak's exception-to-Sentry logging path and existing test setup. Add tests that deliberately trigger the relevant failures, inspect records sent to the configured Sentry DSN, and verify production execution does not crash.

Written by the indexing model from the issue text.