Redis Cache Invalidation LAN Local Area Network Diagram

Tech architecture infographic titled "Redis Cache Invalidation" using HOW-IT-WORKS archetype adapted as a cache state machine / lifecycle. Show a blueprint schematic of a Redis-backed application on a local area network style layout, but keep all on-image wording strictly focused on software architecture. Include labeled boxes connected by directional arrows: Browser — "User client sending HTTPS requests"; API Service — "REST application handling reads and writes"; Redis Cache — "In-memory key-value store for hot data"; Primary Database — "Persistent source of truth for records"; Message Queue — "Async invalidation event transport"; Worker / Consumer — "Processes invalidation events and refresh tasks"; Generic Cloud / LAN Network — "Private network path between services". Add optional state boxes for cache lifecycle: Cache Miss — "Key not found in Redis"; Cache Hit — "Fresh value served from cache"; Stale Entry — "Cached value outdated after data change"; Invalidated Key — "Entry deleted or version bumped"; Rehydrated Cache — "Fresh value repopulated from database". Connect with arrows and short English labels showing exact data flow: Browser to API "HTTPS GET /resource"; API to Redis Cache "GET key"; Redis Cache to API "value / nil"; API to Database "SELECT row"; Database to API "row data"; API to Redis Cache "SET key EX ttl"; API to Browser "200 OK JSON"; Browser to API "HTTPS PUT /resource"; API to Database "UPDATE row"; Database to API "commit OK"; API to Message Queue "publish invalidation event"; Message Queue to Worker / Consumer "key, entity ID, version"; Worker / Consumer to Redis Cache "DEL key" or "UNLINK key"; Worker / Consumer to Redis Cache "SET refreshed value"; API to Redis Cache "bypass stale key"; Redis Cache to API "fresh value". Make the lifecycle visually explicit with numbered stages and state transitions: read request enters, cache miss path, database fetch, cache population, subsequent cache hit, write/update event, invalidation, next read refresh. Add small status/code annotations where accurate: "200 OK", "404 Not Found" optional for absent DB row, "204 No Content" optional for delete acknowledgment. Mention Redis-specific invalidation patterns in callouts: "TTL expiration", "write-through", "cache-aside", "event-driven invalidation", "versioned keys", "pub/sub notification". Keep protocol labels technically accurate: HTTPS between Browser and API, TCP between services where shown, Redis protocol implied for cache commands, SQL or query channel for DB access. Include a numbered legend 1-7 in English: 1. Client sends HTTPS read request to API. 2. API checks Redis for key and receives hit or miss. 3. On miss, API queries primary database for the current record. 4. API returns 200 OK JSON and stores fresh value in Redis with TTL. 5. A later write updates the database as source of truth. 6. API or worker publishes and processes invalidation event, deleting or versioning stale cache keys. 7. Next read repopulates Redis with fresh data and resumes cache-hit path. Visual style: editorial developer-blog illustration, isometric or flat tech-diagram style, vector-clean infographic layout, blueprint schematic, architect-grade clarity, precise linework, minimal monochrome palette with deep navy background, white and light gray strokes, subtle cyan highlights for active arrows, restrained technical mood, no marketing gloss, no audit or security certification cues. Use gridded background, fine measurement marks, clean box hierarchy, compact captions, and LAN-like topology spacing suitable for staff-level review. All text MUST be written in English (array). Every heading, label, caption, legend and metric name in the image must be in English — not English. Spell each English word correctly using English characters and diacritics. Numbers stay as digits, no real cloud-vendor logos (AWS / GCP / Azure) — use generic cloud icons, no watermarks No real cloud-vendor logos (AWS, GCP, Azure) beyond generic cloud icons. Common protocol names (HTTPS, TCP, JWT, OAuth, REST, GraphQL) stay in canonical English form. No security-claim overstatements (do not present diagrams as audited reference architectures).