Break things on purpose - Chaos testing with NResilience
Stop hoping your resilience policies work. Use the Chaos tool to inject faults through the pipeline and verify your system's behavior under pressure.
Read the postWriting
Building a resilience library in the open
Deep dives into the architecture of NResilience - single-loop execution, memory budgets, and the low-level trade-offs of building a high-performance resilience engine for .NET.
Scripts, not anecdotes - Deep dive into Simulation
Stop guessing how your system recovers from an outage. Use scripted simulations to turn resilience behavior into a deterministic, assertable fact.
Stress-tested against Polly
I ran NResilience and Polly 8 through a sustained-load stress suite for 32 minutes. Here are the numbers.
A deep dive into the circuit breaker
The breaker state machine, adaptive trips that measure instead of guess, the 100ms guarded rejection, the recovery ramp, and equal jitter - how each piece works and why it is there.
A deep dive into retries and the retry budget
Why attempt limits cannot stop a retry storm, how a token bucket funded by successful traffic can, the two things it exempts, the hint it refuses to overstate, and the nested-retry marker for the loops it cannot see.
Beyond the basics: The road to v1.0
A final synthesis of the NResilience series, reflecting on the journey from a single retried call to a high-performance resilience engine.
Shipping a resilience library
A meta-look at the quality gates, analyzer bundling, and CI processes that ensure NResilience's promises of performance and correctness.
The small decisions that add up
An essay on the deliberate design choices in NResilience, from guarded rejections to live quantiles and the anatomy of a Verdict.
One loop, one allocation
A deep dive into the NResilience flat executor architecture, its allocation budgets, and why a single loop is faster than a layered pipeline.
gRPC in depth
Explore how NResilience handles gRPC status classification, per-service scoping, and the critical math behind wire deadlines.
HTTP in depth
Learn how NResilience handles per-host state, transport timeouts, wire idempotency, and nested retry detection in .NET HTTP clients.
Tests that don't flake
How to write deterministic resilience tests using scripted callbacks, fake clocks, and event recording.
Deadlines that travel
How to propagate deadlines and criticality across services in ASP.NET Core, avoiding the nested-retry storm and the "zombie request" problem.
Page 1 of 3