Sentry's SDK hooks into your runtime environment and automatically reports errors, uncaught exceptions, and unhandled rejections as well as other types of errors depending on the platform.
- An event is one instance of sending data to Sentry. Generally, this data is an error or exception.
- An issue is a grouping of similar events.
- The reporting of an event is called capturing. When an event is captured, it’s sent to Sentry.
The most common form of capturing is to capture errors. What can be captured as an error varies by platform. In general, if you have something that looks like an exception, it can be captured. For some SDKs, you can also omit the argument to
capture_exception and Sentry will attempt to capture the current exception. It is also useful for manual reporting of errors or messages to Sentry.
While capturing an event, you can also record the breadcrumbs that lead up to that event. Breadcrumbs are different from events: they will not create an event in Sentry, but will be buffered until the next event is sent. Learn more about breadcrumbs in our Breadcrumbs documentation.
In Python you can either capture a caught exception or the one currently held in
sys.exc_info() by not passing an argument:
from sentry_sdk import capture_exception try: a_potentially_failing_function() except Exception as e: # Alternatively the argument can be omitted capture_exception(e)
Another common operation is to capture a bare message. A message is textual information that should be sent to Sentry. Typically messages are not emitted, but they can be useful for some teams.
from sentry_sdk import capture_message capture_message('Something went wrong')
Our documentation is open source and available on GitHub. Your contributions are welcome, whether fixing a typo (drat!) to suggesting an update ("yeah, this would be better").