Secondary storage mechanism strategy for job creation/processing failed due to connection issues

Our app is hosted in azure, sometimes there connection level issues to the redis instances, is there some way to hook into the job creation logic and if for some reason there is a connection issue, the job can be stuffed into another data source, and then we could configure a separate worker to always check that data source and if found it could just requeue them into the correct queue once the connection is re-established.