OFFSET gets slower as tables grow and becomes inconsistent under writes. Keyset pagination is stable and fast: paginate by (created_at, id) cursor. This is a common “senior Rails” upgrade for activity feeds.

Most ‘caching’ bugs are really invalidation bugs, so I stick to a simple cache-aside pattern with conservative TTLs and treat cache misses as normal. The big failure mode to avoid is a stampede: if many requests miss at once, you can crush your DB. Fo

Backfills often fail because we accidentally load full records and associations. Use select to fetch only needed columns and find_each to keep memory flat. This is basic, but it’s where outages come from.

Heavy work inside a transaction increases lock time and deadlock risk. Use after_save_commit to schedule slow tasks (thumb generation, external sync) once the write is durable.

Background jobs will fail in production, so I like having a predictable story for retries and poison messages. BullMQ is a solid middle ground: Redis-backed, straightforward, and good enough for most apps. I set explicit attempts and backoff, and when

ETags are a simple high-impact performance lever: they reduce bandwidth and controller CPU. Use stale? with a stable ETag component list and last_modified so intermediary caches and browsers can do conditional GET.

Even without extra gems, you can validate incoming JSON payloads with small “contracts” that coerce and validate keys. It’s a strong reliability upgrade for webhook and API ingestion.

Direct uploads are great because they keep file traffic away from your Rails dynos, but the default UX is opaque. I attach a Stimulus controller that listens for Active Storage’s direct-upload:* events and updates a progress bar. This keeps the markup

Modals are accessibility traps—literally. Without focus management, keyboard users can tab into the page behind the modal and get lost. I trap focus within the dialog while it’s open, restore focus to the trigger on close, and support Escape to dismis

If you ever need to hand logs to support, you don’t want secrets in them. Filter params at the framework level; then add custom filters for app-specific fields (API keys, tokens).

Rails enums are nice, but the DB should enforce allowed values. Use a CHECK constraint (or native enum type) plus the Rails enum mapping. It prevents bad writes from console scripts and future migrations.

Streaming responses serve large files without loading them entirely into memory. I use StreamingHttpResponse or FileResponse for file downloads. For CSV generation, I yield rows incrementally. The generator pattern keeps memory usage constant regardle