Find agent skills by outcome

- Hot key mitigation: Local caching of hot keys

key replication across shards

- Cache-aside (lazy loading): Application manages cache reads and writes explicitly

simple but risk of stale reads

- Write-behind (write-back): Asynchronous batch writes to backing store

lower write latency

- Write-through: Synchronous cache update on every write

strong consistency

- Database caching: Query result caching

materialized views

- Application-level caching: In-process caches (e.g.

Guava

- Distributed caching: Redis clusters for shared state

Memcached for simple key-value hot data

- Plan for cache stampede prevention using locking

probabilistic early expiration

- Implement cache hit rate monitoring

latency tracking

- CDN caching: Static assets

dynamic page caching with edge-side includes

### 5. Monitoring

Failover

- Configure eviction policies (LRU

LFU

- Calculate cache memory requirements based on data size

cardinality

- Select appropriate caching technologies (Redis

Memcached

- Select invalidation patterns per data type: write-through for critical data

write-behind for write-heavy workloads

Creative

- Design from the outside in: CDN layer

application cache layer

Creative

- Select optimal caching technologies based on use case requirements

designing hybrid solutions that combine multiple technologies including CDN and edge caching

- Architect distributed caching solutions that scale horizontally with cache warming

preloading

- Identify hot data sets

access patterns

- Optimize cache hit rates through strategic cache placement

sizing

- Implement cache invalidation patterns including write-through

write-behind

- Ensure data consistency by designing invalidation workflows

eventual consistency patterns

Creative

- Design multi-layer caching architectures using Redis

Memcached

Productivity

- Assign each task a stable ID (e.g.

TASK-1.1) and use checklist items in outputs.