Optimizing performance

Monitor and Troubleshoot Azure Solutions

Ebadur Osib

Senior Cloud Consulting Engineer

Performance is a feature

Performance impacts user trust.
Slow apps increase failures.
Azure helps developers optimize for:
- Low latency
- High traffic

Where performance problems come from

Latency vs throughput

Optimize the critical path

Optimization starts by identifying and shortening the slowest sequence.
Not everything requires equal tuning.

Reducing dependency latency

Dependencies are the most common source of latency.
Methods for reduction:
- Query optimization
- Connection pooling
- Batching requests
- Reducing chatty calls
More effective than scaling compute.

Caching

Azure Cache for Redis

¹ https://learn.microsoft.com/en-us/azure/redis/architecture

Azure Managed Redis

¹ https://learn.microsoft.com/en-us/azure/redis/architecture

Cache design considerations

Caching introduces important design decisions.
You must define expiration policies for cached data.
Trade-offs exist between consistency and freshness.
A common pattern is cache-aside, where data is loaded into the cache only when needed.

Cache invalidation ensures outdated data is refreshed or removed.

Throttling and backpressure

Too many requests can cause failures.
Throttling and backpressure protect systems.
Slows or rejects excess traffic.

Asynchronous and queue-based patterns

The business problem

Finding the bottleneck

Optimizing for scale and speed

Let's practice!

Monitor and Troubleshoot Azure Solutions

Preparing Video For Download...