Redis vs Database Queue Driver for Sub‑100 ms Job Latency in Lumen

Redis vs Database Queue Driver for Sub‑100 ms Job Latency in Lumen

Goal: Ensure Lumen queue processing stays under 100 ms during peak traffic while preserving job durability and avoiding worker restarts after changing the queue connection.

Redis offers in‑memory speed but requires persistence configuration to prevent loss on restart; the database driver is durable but slower and does not automatically retry failed jobs.

Which driver can reliably meet the sub‑100 ms latency requirement under peak load? Does the database driver support automatic retries for failed jobs? Can queue connection changes be applied without restarting workers?