You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There is a nice hook for handling exceptions raised by blocks registered with every, but no such hook for instrumenting the execution of those blocks.
What happened for us was that a block was failing to execute and we were not aware since we did not register an error_handler. This worked for reporting errors, but did not include enough context to tell us that the error came from clockwork, which is required to implement a different alert policy on failure (create an incident after a very small number of failures, as opposed to relying on the error rate of our already high-throughput web application to create an incident).
What we ended up doing was patching the every method to instrument execution of the registered block with the New Relic Ruby gem, as such:
There is a nice hook for handling exceptions raised by blocks registered with
every
, but no such hook for instrumenting the execution of those blocks.What happened for us was that a block was failing to execute and we were not aware since we did not register an
error_handler
. This worked for reporting errors, but did not include enough context to tell us that the error came from clockwork, which is required to implement a different alert policy on failure (create an incident after a very small number of failures, as opposed to relying on the error rate of our already high-throughput web application to create an incident).What we ended up doing was patching the
every
method to instrument execution of the registered block with the New Relic Ruby gem, as such:This seems like behavior that could be added to the
Manager
in a clean way. Thoughts?The text was updated successfully, but these errors were encountered: