AZ-305 ยท Application architecture design
Application caching: Azure Managed Redis vs edge caching
Azure Managed Redis and Azure Cache for Redis use cases (cache-aside, session store), tiers and clustering, versus CDN or Front Door edge caching for static content.
Key points
- Azure Managed Redis: in-memory store, successor to Azure Cache for Redis.
- Patterns: cache-aside (load on request, with expiry), session store, leaderboards (sorted sets).
- Tiers: Memory Optimized 8:1, Balanced 4:1, Compute Optimized 2:1, Flash Optimized (cold data on NVMe).
- HA: shards on 2+ nodes; active geo-replication on in-memory tiers.
- Edge caching (Front Door, CDN): static files near users, not app data.
Source: https://learn.microsoft.com/en-us/azure/redis/overview
Practice questions
Which caching pattern loads an item into the cache only when the application first requests it and does not find it there?
- Write-through
- Cache-aside
- Refresh-ahead
- Session store
Show answer
Answer: B. Cache-aside
Cache-aside: the application checks the cache, and on a miss reads the data store and populates the cache. Write-through writes on every update, refresh-ahead preloads before expiry, and a session store is a use case, not a loading pattern.
Which workloads are a good fit for Azure Managed Redis rather than an edge cache such as Azure Front Door?
- Shopping-cart session state for a web farm
- Results of a slow product-price database query
- A real-time game leaderboard
- Product images served to users worldwide
- Large installer downloads for customers
Show answer
Answer: A. Shopping-cart session state for a web farm, B. Results of a slow product-price database query, C. A real-time game leaderboard
Redis keeps application data in memory: user sessions, results of expensive database queries, and leaderboards built on sorted sets. Static images and downloadable files for a global audience belong at the edge, cached by Front Door or a CDN close to the users.
Choosing the Flash Optimized tier of Azure Managed Redis makes the cached data more resilient, because part of it is stored on disk.
Show answer
Answer: False
False. Flash Optimized stores cold data on NVMe to cut cost for large datasets; it does not increase data resiliency. For recovery after an outage you enable data persistence (available on all tiers) and high availability.
Practice 8 questions on this topic in Bitzu
Short lessons, instant explanations and spaced review of what you get wrong. Free on iOS and Android.
