Skip to content

Conversation

@nickvergessen
Copy link
Member

Backport of #758

So before this PR, we tried to act on 400 errors already,
but I didn't know that they throw an exception on the `post()`,
so instead of performing on the response body,
we `continue;` after we logged an info hushed

This is of course not what was intended. Further more,
since the notifications are grouped and the proxy had an error
which would immediately bail out with a 400 when there was 1 broken notification,
instead of sending the rest, 1 broken notification was enough to not
send any notification anymore. And since we never deleted the broken
device the instance didn't "heal" itself, but always broke like this.

Signed-off-by: Joas Schilling <[email protected]>
@rullzer rullzer mentioned this pull request Oct 1, 2020
1 task
@nickvergessen nickvergessen merged commit 965a8fa into stable18 Oct 1, 2020
@nickvergessen nickvergessen deleted the backport/758/stable18 branch October 1, 2020 13:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants