Classes which implement the perform() method can now also implement
before_perform() and after_perform() methods which take a metadata dict.
These contain the args passed in, as well as timestamps for when the job
was enqueued and when it was performed, and whether the job failed and
was retried.
I'm using it mainly so I can track queue processing latency, and the amount of time it takes to process each job. I send them along to a statsd/graphite server.
Classes which implement the perform() method can now also implement before_perform() and after_perform() methods which take a metadata dict. These contain the args passed in, as well as timestamps for when the job was enqueued and when it was performed, and whether the job failed and was retried.
I'm using it mainly so I can track queue processing latency, and the amount of time it takes to process each job. I send them along to a statsd/graphite server.