Redis Health Check

Pings Redis and evaluates responsiveness. Also collects useful metrics (clients, memory, ops/sec, etc.).

Install

dotnet add package Monivus.HealthChecks.Redis

Usage

using Monivus.HealthChecks;

builder.Services.AddHealthChecks()
    .AddRedisEntry();

If not providing a connection string in configuration, ensure an IConnectionMultiplexer is registered.

Configuration

Binding path: Monivus:Redis

ConnectionString

Gets or sets the Redis connection string.

{
  "Monivus": {
    "Redis": { 
        "ConnectionString": "localhost:6379" 
        } 
    }
}

SlowPingThresholdMilliseconds

Threshold in milliseconds for considering a Redis PING as slow.

Status rules:

• Healthy when the PING completes and (if set) the latency is less than or equal to the threshold.
• Degraded when the measured PING latency is greater than the threshold.
• Unhealthy when the connection is not established, there are no endpoints, or exceptions occur.

{
  "Monivus": {
    "Redis": {
      "ConnectionString": "localhost:6379",
      "SlowPingThresholdMilliseconds": 200
    }
  }
}

Example interpretation: if a PING takes above 200 ms, the check reports Degraded; at or below 200 ms it’s Healthy. Connection or command errors result in Unhealthy.

Admin Mode (AllowAdmin)

Some extended metrics are retrieved via the Redis INFO command. In StackExchange.Redis this is treated as a server/admin operation. If the connection has admin operations disabled (AllowAdmin = false, which is the default) or the command is unavailable, attempting to call server.InfoAsync() may throw.

This health check handles that gracefully:

• It still performs a PING and basic probes (e.g., server info, DB size, last save) and reports Healthy/Degraded based on latency and connectivity.
• When admin is not allowed, it returns Healthy with limited data and the description includes a hint: Redis is healthy and responsive. (Admin not allowed!).
• Extended metrics (memory usage breakdown, connected clients, ops/sec, hit/miss, etc.) are omitted in this mode.

To enable extended metrics, allow admin commands for the connection used by the health check:

# Connection string option
localhost:6379,allowAdmin=true

// ConfigurationOptions example
var opts = ConfigurationOptions.Parse("localhost:6379");
opts.AllowAdmin = true;
var mux = await ConnectionMultiplexer.ConnectAsync(opts);

builder.Services.AddSingleton<IConnectionMultiplexer>(mux);
builder.Services.AddHealthChecks().AddRedisEntry("Redis");

Tip: If you prefer to keep application connections non-admin, you can maintain a separate admin-enabled multiplexer registered only for health/metrics.

Example Entries (JSON)

Below are example payloads for entries.Redis as returned by the /health endpoint.

Healthy with admin enabled (full metrics):

{
  "status": "Healthy",
  "description": "Redis is healthy and responsive.",
  "duration": "00:00:00.0030000",
  "durationMs": 3,
  "data": {
    "IsConnected": true,
    "ServerVersion": "7.2.4",
    "ServerType": "Standalone",
    "PingMilliseconds": 12,
    "DatabaseSize": 2048,
    "LastSaveUtc": "2024-06-12T08:21:45.0000000Z",
    "UsedMemoryMb": 128.5,
    "UsedMemoryRssMb": 140.1,
    "TotalSystemMemoryMb": 16384,
    "MemoryUsagePercent": 0.78,
    "ConnectedClients": 5,
    "BlockedClients": 0,
    "OpsPerSecond": 342,
    "UptimeSeconds": 86400,
    "MemoryFragmentationRatio": 1.07,
    "KeyspaceHits": 15000,
    "KeyspaceMisses": 500,
    "KeyspaceHitRatePercent": 96.77
  },
  "exception": null,
  "tags": ["Redis"],
  "entryType": "Redis"
}

Healthy with admin disabled (limited metrics):

{
  "status": "Healthy",
  "description": "Redis is healthy and responsive. (Admin not allowed!)",
  "duration": "00:00:00.0020000",
  "durationMs": 2,
  "data": {
    "IsConnected": true,
    "ServerVersion": "7.2.4",
    "ServerType": "Standalone",
    "PingMilliseconds": 10,
    "DatabaseSize": 1024,
    "LastSaveUtc": "2024-06-12T08:21:45.0000000Z"
  },
  "exception": null,
  "tags": ["Redis"],
  "entryType": "Redis"
}

Degraded when ping exceeds threshold (no extended metrics, matches current implementation order):

{
  "status": "Degraded",
  "description": "Redis ping exceeded threshold (210ms).",
  "duration": "00:00:00.0040000",
  "durationMs": 4,
  "data": {
    "IsConnected": true,
    "ServerVersion": "7.2.4",
    "ServerType": "Standalone",
    "PingMilliseconds": 210,
    "DatabaseSize": 1024,
    "LastSaveUtc": "2024-06-12T08:21:45.0000000Z"
  },
  "exception": null,
  "tags": ["Redis"],
  "entryType": "Redis"
}