Skip to content

Conversation

@nateberkopec
Copy link
Contributor

@nateberkopec nateberkopec commented Sep 20, 2016

Note this won't catch exceptions in threads, obviously.

See also #511 and #506


This change is Reviewable

Note this won't catch exceptions in threads, obviously.
@nateberkopec
Copy link
Contributor Author

Anyone know of any jiu-jitsu I can do if someone is using a Thread that raises an exception?

@nuclearpidgeon
Copy link

You could perhaps wrap the exception handling code in a lambda/block, and pass that as another argument to the config.async lambda/block - but it's pretty superfluous at that point as that handler would just call 'send_event' anyway.. which presumably would have just failed for some reason. Unless you want to just pass in the Logger call.

Also just a heads up, that config.rst page doesn't seem to render properly in GitHub. Not sure if this is just a GitHub problem or some kind of syntax error

@nateberkopec
Copy link
Contributor Author

Hmm. We could call the transport_failure_callback.. Normally that's called when there's a 500 error from Sentry, but it might be appropriate here too.

@dcramer
Copy link
Member

dcramer commented Sep 21, 2016

Of note 429 and 403 codes, for example, should not be retried. So as long as it's only 500 it might be ok. That's still not great behavior though as it means sdks could hammer sentry in certain situations.

@nateberkopec
Copy link
Contributor Author

Actually, send_event will just call transport_failure_callback anyway if Sentry returns an error. I think a loud (error-level) log message is probably enough here.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants