Simmy docs

.github/wordlist.txt

@@ -32,6 +32,7 @@ nuget
 oss
 pcl
 parallelize
+paas
 pluralsight
 pollydocs
 pre
@@ -47,6 +48,7 @@ rethrow
 rethrows
 retryable
 reusability
+saas
 sdk
 serializers
 silverlight
@@ -60,6 +62,7 @@ telemetrylistener
 testability
 timingpolicy
 ui
+unhandled
 uwp
 waitandretry
 wpf

docs/chaos/behavior.md

@@ -0,0 +1,134 @@
+# Behavior monkey strategy
+
+## About
+
+- **Options**: [`BehaviorStrategyOptions`](xref:Polly.Simmy.Behavior.BehaviorStrategyOptions)
+- **Extensions**: `AddChaosBehavior`
+- **Strategy Type**: Proactive
+
+---
+
+The behavior chaos strategy is designed to inject custom behaviors into system operations right before such an operation is invoked. This strategy is flexible, allowing users to define specific behaviors such as altering the input, simulating resource exhaustion, putting the system in a given state before the actual operation is called, or other operational variations to simulate real-world scenarios.
+
+## Usage
+
+<!-- snippet: chaos-behavior-usage -->
+```cs
+// Behavior using the default options.
+// See https://www.pollydocs.org/chaos/behavior#defaults for defaults.
+var optionsDefault = new BehaviorStrategyOptions();
+
+// To use a custom function to generate the behavior to inject.
+var optionsWithBehaviorGenerator = new BehaviorStrategyOptions
+{
+    BehaviorAction = static args => RestartRedisVM(),
+    Enabled = true,
+    InjectionRate = 0.05
+};
+
+// To get notifications when a behavior is injected
+var optionsOnBehaviorInjected = new BehaviorStrategyOptions
+{
+    BehaviorAction = static args => RestartRedisVM(),
+    Enabled = true,
+    InjectionRate = 0.05,
+    OnBehaviorInjected = static args =>
+    {
+        Console.WriteLine("OnBehaviorInjected, Operation: {0}.", args.Context.OperationKey);
+        return default;
+    }
+};
+
+// Add a behavior strategy with a BehaviorStrategyOptions instance to the pipeline
+new ResiliencePipelineBuilder().AddChaosBehavior(optionsDefault);
+new ResiliencePipelineBuilder<HttpStatusCode>().AddChaosBehavior(optionsWithBehaviorGenerator);
+
+// There are also a handy overload to inject the chaos easily.
+new ResiliencePipelineBuilder().AddChaosBehavior(0.05, RestartRedisVM);
+```
+<!-- endSnippet -->
+
+Example execution:
+
+<!-- snippet: chaos-behavior-execution -->
+```cs
+var pipeline = new ResiliencePipelineBuilder()
+    .AddRetry(new RetryStrategyOptions
+    {
+        ShouldHandle = new PredicateBuilder().Handle<RedisConnectionException>(),
+        BackoffType = DelayBackoffType.Exponential,
+        UseJitter = true,  // Adds a random factor to the delay
+        MaxRetryAttempts = 4,
+        Delay = TimeSpan.FromSeconds(3),
+    })
+    .AddChaosBehavior(new BehaviorStrategyOptions // Monkey strategies are usually placed as the last ones in the pipeline
+    {
+        BehaviorAction = static args => RestartRedisVM(),
+        Enabled = true,
+        InjectionRate = 0.05
+    })
+    .Build();
+```
+<!-- endSnippet -->
+
+## Defaults
+
+| Property             | Default Value | Description                                    |
+|----------------------|---------------|------------------------------------------------|
+| `OnBehaviorInjected` | `null`        | Action executed when the behavior is injected. |
+| `BehaviorAction`     | `null`        | Custom behavior to be injected.                |
+
+## Diagrams
+
+### Normal 🐵 sequence diagram
+
+```mermaid
+sequenceDiagram
+    actor C as Caller
+    participant P as Pipeline
+    participant B as Behavior
+    participant D as DecoratedUserCallback
+
+    C->>P: Calls ExecuteAsync
+    P->>B: Calls ExecuteCore
+    activate B
+    B-->>B: Determines Injection<br/>Decision: 🐵
+    deactivate B
+    B->>+D: Invokes
+    D->>-B: Returns result
+    B->>P: Returns result
+    P->>C: Returns result
+```
+
+### Chaos 🙈 sequence diagram
+
+```mermaid
+sequenceDiagram
+    actor C as Caller
+    participant P as Pipeline
+    participant B as Behavior
+    participant D as DecoratedUserCallback
+
+    C->>P: Calls ExecuteAsync
+    P->>B: Calls ExecuteCore
+    activate B
+    B-->>B: Determines Injection<br/>Decision: 🙈
+    B-->>B: Injects Behavior
+    deactivate B
+    B->>+D: Invokes
+    D->>-B: Returns result
+    B->>P: Returns result
+    P->>C: Returns result
+```
+
+## Anti-patterns
+
+### Injecting delay
+
+❌ DON'T
+
+Use behavior strategies to inject delays.
+
+✅ DO
+
+Use the latency monkey instead as the [`LatencyChaosStrategy`](latency.md) already correctly handles synchronous/asynchronous delay executions, cancellations, etc.

docs/chaos/fault.md

@@ -0,0 +1,140 @@
+# Fault monkey strategy
+
+## About
+
+- **Options**: [`FaultStrategyOptions`](xref:Polly.Simmy.Fault.FaultStrategyOptions)
+- **Extensions**: `AddChaosFault`
+- **Strategy Type**: Proactive
+
+---
+
+The fault chaos strategy is designed to introduce faults (exceptions) into the system, simulating real-world scenarios where operations might fail unexpectedly. It is configurable to inject specific types of exceptions or use custom logic to generate faults dynamically.
+
+## Usage
+
+<!-- snippet: chaos-fault-usage -->
+```cs
+// Fault using the default options.
+// See https://www.pollydocs.org/chaos/fault#defaults for defaults.
+var optionsDefault = new FaultStrategyOptions();
+
+// 10% of invocations will be randomly affected.
+var basicOptions = new FaultStrategyOptions
+{
+    Fault = new InvalidOperationException("Dummy exception"),
+    Enabled = true,
+    InjectionRate = 0.1
+};
+
+// To use a custom function to generate the fault to inject.
+var optionsWithFaultGenerator = new FaultStrategyOptions
+{
+    FaultGenerator = static args =>
+    {
+        Exception? exception = args.Context.OperationKey switch
+        {
+            "DataLayer" => new TimeoutException(),
+            "ApplicationLayer" => new InvalidOperationException(),
+            _ => null // When the fault generator returns null the strategy won't inject any fault and it will just invoke the user's callback
+        };
+
+        return new ValueTask<Exception?>(exception);
+    },
+    Enabled = true,
+    InjectionRate = 0.1
+};
+
+// To get notifications when a fault is injected
+var optionsOnFaultInjected = new FaultStrategyOptions
+{
+    Fault = new InvalidOperationException("Dummy exception"),
+    Enabled = true,
+    InjectionRate = 0.1,
+    OnFaultInjected = static args =>
+    {
+        Console.WriteLine("OnFaultInjected, Exception: {0}, Operation: {1}.", args.Fault.Message, args.Context.OperationKey);
+        return default;
+    }
+};
+
+// Add a fault strategy with a FaultStrategyOptions instance to the pipeline
+new ResiliencePipelineBuilder().AddChaosFault(optionsDefault);
+new ResiliencePipelineBuilder<HttpStatusCode>().AddChaosFault(optionsWithFaultGenerator);
+
+// There are also a couple of handy overloads to inject the chaos easily.
+new ResiliencePipelineBuilder().AddChaosFault(0.1, () => new InvalidOperationException("Dummy exception"));
+```
+<!-- endSnippet -->
+
+Example execution:
+
+<!-- snippet: chaos-fault-execution -->
+```cs
+var pipeline = new ResiliencePipelineBuilder()
+    .AddRetry(new RetryStrategyOptions
+    {
+        ShouldHandle = new PredicateBuilder().Handle<InvalidOperationException>(),
+        BackoffType = DelayBackoffType.Exponential,
+        UseJitter = true,  // Adds a random factor to the delay
+        MaxRetryAttempts = 4,
+        Delay = TimeSpan.FromSeconds(3),
+    })
+    .AddChaosFault(new FaultStrategyOptions // Monkey strategies are usually placed as the last ones in the pipeline
+    {
+        Fault = new InvalidOperationException("Dummy exception"),
+        Enabled = true,
+        InjectionRate = 0.1
+    })
+    .Build();
+```
+<!-- endSnippet -->
+
+## Defaults
+
+| Property          | Default Value | Description                                          |
+|-------------------|---------------|------------------------------------------------------|
+| `OnFaultInjected` | `null`        | Action executed when the fault is injected.          |
+| `FaultGenerator`  | `null`        | Generates the fault to inject for a given execution. |
+| `Fault`           | `null`        | The fault to inject.                                 |
+
+## Diagrams
+
+### Normal 🐵 sequence diagram
+
+```mermaid
+sequenceDiagram
+    actor C as Caller
+    participant P as Pipeline
+    participant F as Fault
+    participant D as DecoratedUserCallback
+
+    C->>P: Calls ExecuteAsync
+    P->>F: Calls ExecuteCore
+    activate F
+    F-->>F: Determines Injection<br/>Decision: 🐵
+    deactivate F
+    F->>+D: Invokes
+    D->>-F: Returns result
+    F->>P: Returns result
+    P->>C: Returns result
+```
+
+### Chaos 🙈 sequence diagram
+
+```mermaid
+sequenceDiagram
+    actor C as Caller
+    participant P as Pipeline
+    participant F as Fault
+    participant D as DecoratedUserCallback
+
+    C->>P: Calls ExecuteAsync
+    P->>F: Calls ExecuteCore
+    activate F
+    F-->>F: Determines Injection<br/>Decision: 🙈
+    F-->>F: Injects Fault
+    deactivate F
+    Note over D: The user's Callback is not invoked<br/>when a fault is injected
+    F->>P: Throws injected Fault
+    P->>C: Propagates Exception
+```

docs/chaos/index.md

@@ -0,0 +1,53 @@
+# Chaos engineering with Simmy
+
+[Simmy][simmy] is a major new companion project adding a chaos-engineering and fault-injection dimension to Polly, through the provision of strategies to selectively inject faults, latency, custom behavior or fake results.
+
+![Simmy](../media/simmy-logo.png)
+
+## Motivation
+
+There are a lot of questions when it comes to chaos-engineering and making sure that a system is actually ready to face the worst possible scenarios:
+
+* Is my system resilient enough?
+* Am I handling the right exceptions/scenarios?
+* How will my system behave if X happens?
+* How can I test without waiting for a handled (or even unhandled) exception to happen in my production environment?
+
+Using Polly helps introduce resilience to a project, but we don't want to have to wait for expected or unexpected failures to test it out. A resilience could be wrongly implemented; testing the scenarios is not straightforward; and mocking failure of some dependencies (for example a cloud SaaS or PaaS service) is not always straightforward.
+
+### What is needed to simulate chaotic scenarios?
+
+* A way to simulate failures of dependencies (any service dependency for example).
+* Define when to fail based on some external factors - maybe global configuration or some rule.
+* A way to revert easily, to control the blast radius.
+* To be production grade, to run this in a production or near-production system with automation.
+
+## Chaos strategies (a.k.a Monkey strategies)
+
+Chaos strategies (or Monkey strategies as we call them) are in essence a [Resilience strategy](../strategies/index.md#built-in-strategies), which means, as a *Resilience Strategy* is the minimum unit of resilience for Polly, a *Monkey Strategy* is the minimum unit of chaos for Simmy.
+
+### Built-in strategies
+
+| Strategy                | Reactive | What does the strategy do?                                           |
+|-------------------------|----------|----------------------------------------------------------------------|
+| [Fault](fault.md)       | No       | Injects exceptions in your system.                                   |
+| [Result](result.md)     | Yes      | Substitute results to fake outcomes in your system.                  |
+| [Latency](latency.md)   | No       | Injects latency into executions before the calls are made.           |
+| [Behavior](behavior.md) | No       | Allows you to inject *any* extra behaviour, before a call is placed. |
+
+## Usage
+
+It is usual to place the monkey strategy as the last strategy in the resilience pipeline. By placing the monkey strategies as last, they subvert the usual outbound call at the last minute, substituting their fault or adding extra latency, etc. The existing resilience strategies - further out in the `ResiliencePipeline` - still apply, so you can test how the Polly resilience strategies you have configured handle the chaos/faults injected by Simmy.
+
+## Common options across strategies
+
+All the strategies' options implement the [`MonkeyStrategyOptions`](xref:Polly.Simmy.MonkeyStrategyOptions) class as it contains the basic configuration for every monkey strategy.
+
+| Property                 | Default Value | Description                                                                                                                                                                                                                      |
+|--------------------------|---------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| `InjectionRate`          | 0.001 ms      | A decimal between 0 and 1 inclusive. The strategy will inject the chaos, randomly, that proportion of the time, e.g.: if 0.2, twenty percent of calls will be randomly affected; if 0.01, one percent of calls; if 1, all calls. |
+| `InjectionRateGenerator` | `null`        | Generates the injection rate for a given execution, which the value should be between [0, 1] (inclusive).                                                                                                                        |
+| `Enabled`                | `false`       | Determines whether the strategy is enabled or not.                                                                                                                                                                               |
+| `EnabledGenerator`       | `null`        | the enable generator that indicates whether or not the chaos strategy is enabled for a given execution.                                                                                                                          |
+
+[simmy]: https://github.com/Polly-Contrib/Simmy