Every job has a status. Currently there is a sorted set for each status, where all jobs that have that status are stored. Technically only the "queued" set needs to have a score (the priority) since that allows higher-priority jobs to get picked up first. Using sets instead of sorted sets when possible should reduce memory usage.
Every job has a status. Currently there is a sorted set for each status, where all jobs that have that status are stored. Technically only the "queued" set needs to have a score (the priority) since that allows higher-priority jobs to get picked up first. Using sets instead of sorted sets when possible should reduce memory usage.