Skip to content

Redis

To test the Redis Health Check endpoint, use the following curl command:

curl http://localhost:PORT/health

Health Function

The Health function orchestrates the health assessment of the Redis server by invoking the checkRedisHealth function and returning the collected statistics.

Functionality

Check Redis Health: The function pings the Redis server to check its availability and adds the response to the stats map.

  • If the ping fails, it logs the error and terminates the program.
  • If the ping succeeds, it proceeds to retrieve additional information.

Retrieve Redis Information: The function retrieves information about the Redis server, including version, mode, connected clients, memory usage, uptime, etc.

  • If an error occurs during info retrieval, it updates the health message accordingly.

Evaluate Redis Statistics: The function evaluates the collected statistics to identify potential issues and updates the health message accordingly.

  • It checks for high number of connected clients, stale connections, memory usage, recent restart, high idle connections, and high connection pool utilization.

Sample Output

The Health function returns a JSON-like map structure with various keys representing different health metrics and their corresponding values.

{
  "redis_active_connections": "0",
  "redis_connected_clients": "1",
  "redis_hits_connections": "1",
  "redis_idle_connections": "1",
  "redis_max_memory": "0",
  "redis_message": "Redis has been recently restarted",
  "redis_misses_connections": "1",
  "redis_mode": "standalone",
  "redis_ping_response": "PONG",
  "redis_pool_size_percentage": "0.42%",
  "redis_stale_connections": "0",
  "redis_status": "up",
  "redis_timeouts_connections": "0",
  "redis_total_connections": "1",
  "redis_uptime_in_seconds": "55",
  "redis_used_memory": "980704",
  "redis_used_memory_peak": "980704",
  "redis_version": "7.2.4"
}

Serialization/deserialization

The Sample Output is dynamic and unstructured since it depends on the raw map. To make it structurable, it must implement JSON serialization/deserialization or Other serialization/deserialization (e.g, XML serialization/deserialization) to bind it. For example:

  • JSON serialization/deserialization
{
  "redis_health": {
    "status": "up",
    "message": "Redis connection pool utilization is high",
    "stats": {
      "version": "7.0.15",
      "mode": "standalone",
      "connected_clients": "10",
      "memory": {
        "used": {
          "mb": "22.38",
          "gb": "0.02"
        },
        "peak": {
          "mb": "46.57",
          "gb": "0.05"
        },
        "free": {
          "mb": "1130.00",
          "gb": "1.10"
        },
        "percentage": "1.98%"
      },
      "uptime_stats": "6 days, 3 hours, 37 minutes, 20 seconds",
      "uptime": [
        {
          "day": "6"
        },
        {
          "hour": "3"
        },
        {
          "minute": "37"
        },
        {
          "second": "20"
        }
      ],
      "pooling": {
        "figures": {
          "hits": "10",
          "misses": "2",
          "timeouts": "0",
          "total": "4",
          "stale": "9",
          "idle": "5",
          "active": "0",
          "percentage": "62.50%"
        },
        "observed_total": "26"
      }
    }
  }
}
  • XML serialization/deserialization
<?xml version="1.0" encoding="UTF-8"?>
<redis_health>
  <status>up</status>
  <message>Redis connection pool utilization is high</message>
  <stats>
    <version>7.0.15</version>
    <mode>standalone</mode>
    <connected_clients>10</connected_clients>
    <memory>
      <used>
        <mb>22.38</mb>
        <gb>0.02</gb>
      </used>
      <peak>
        <mb>46.57</mb>
        <gb>0.05</gb>
      </peak>
      <free>
        <mb>1130.00</mb>
        <gb>1.10</gb>
      </free>
      <percentage>1.98%</percentage>
    </memory>
    <uptime_stats>6 days, 3 hours, 37 minutes, 20 seconds</uptime_stats>
    <uptime>
      <day>6</day>
      <hour>3</hour>
      <minute>37</minute>
      <second>20</second>
    </uptime>
    <pooling>
      <figures>
        <hits>10</hits>
        <misses>2</misses>
        <timeouts>0</timeouts>
        <total>4</total>
        <stale>9</stale>
        <idle>5</idle>
        <active>0</active>
        <percentage>62.50%</percentage>
      </figures>
      <observed_total>26</observed_total>
    </pooling>
  </stats>
</redis_health>

Code Implementation

func (s *service) Health() map[string]string {
    ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second) // Default is now 5s
    defer cancel()

    stats := make(map[string]string)

    stats = s.checkRedisHealth(ctx, stats)

    return stats
}

func (s *service) checkRedisHealth(ctx context.Context, stats map[string]string) map[string]string {
    pong, err := s.db.Ping(ctx).Result()
    if err != nil {
        log.Fatalf(fmt.Sprintf("db down: %v", err))
    }

    stats["redis_status"] = "up"
    stats["redis_message"] = "It's healthy"
    stats["redis_ping_response"] = pong

    info, err := s.db.Info(ctx).Result()
    if err != nil {
        stats["redis_message"] = fmt.Sprintf("Failed to retrieve Redis info: %v", err)
        return stats
    }

    redisInfo := parseRedisInfo(info)

    poolStats := s.db.PoolStats()

    stats["redis_version"] = redisInfo["redis_version"]
    stats["redis_mode"] = redisInfo["redis_mode"]
    stats["redis_connected_clients"] = redisInfo["connected_clients"]
    stats["redis_used_memory"] = redisInfo["used_memory"]
    stats["redis_used_memory_peak"] = redisInfo["used_memory_peak"]
    stats["redis_uptime_in_seconds"] = redisInfo["uptime_in_seconds"]
    stats["redis_hits_connections"] = strconv.FormatUint(uint64(poolStats.Hits), 10)
    stats["redis_misses_connections"] = strconv.FormatUint(uint64(poolStats.Misses), 10)
    stats["redis_timeouts_connections"] = strconv.FormatUint(uint64(poolStats.Timeouts), 10)
    stats["redis_total_connections"] = strconv.FormatUint(uint64(poolStats.TotalConns), 10)
    stats["redis_idle_connections"] = strconv.FormatUint(uint64(poolStats.IdleConns), 10)
    stats["redis_stale_connections"] = strconv.FormatUint(uint64(poolStats.StaleConns), 10)
    stats["redis_max_memory"] = redisInfo["maxmemory"]

    activeConns := uint64(math.Max(float64(poolStats.TotalConns-poolStats.IdleConns), 0))
    stats["redis_active_connections"] = strconv.FormatUint(activeConns, 10)

    poolSize := s.db.Options().PoolSize
    connectedClients, _ := strconv.Atoi(redisInfo["connected_clients"])
    poolSizePercentage := float64(connectedClients) / float64(poolSize) * 100
    stats["redis_pool_size_percentage"] = fmt.Sprintf("%.2f%%", poolSizePercentage)

    return s.evaluateRedisStats(redisInfo, stats)
}

func (s *service) evaluateRedisStats(redisInfo, stats map[string]string) map[string]string {
    poolSize := s.db.Options().PoolSize
    poolStats := s.db.PoolStats()
    connectedClients, _ := strconv.Atoi(redisInfo["connected_clients"])
    highConnectionThreshold := int(float64(poolSize) * 0.8)

    if connectedClients > highConnectionThreshold {
        stats["redis_message"] = "Redis has a high number of connected clients"
    }

    minStaleConnectionsThreshold := 500
    if int(poolStats.StaleConns) > minStaleConnectionsThreshold {
        stats["redis_message"] = fmt.Sprintf("Redis has %d stale connections.", poolStats.StaleConns)
    }

    usedMemory, _ := strconv.ParseInt(redisInfo["used_memory"], 10, 64)
    maxMemory, _ := strconv.ParseInt(redisInfo["maxmemory"], 10, 64)
    if maxMemory > 0 {
        usedMemoryPercentage := float64(usedMemory) / float64(maxMemory) * 100
        if usedMemoryPercentage >= 90 {
            stats["redis_message"] = "Redis is using a significant amount of memory"
        }
    }

    uptimeInSeconds, _ := strconv.ParseInt(redisInfo["uptime_in_seconds"], 10, 64)
    if uptimeInSeconds < 3600 {
        stats["redis_message"] = "Redis has been recently restarted"
    }

    idleConns := int(poolStats.IdleConns)
    highIdleConnectionThreshold := int(float64(poolSize) * 0.7)
    if idleConns > highIdleConnectionThreshold {
        stats["redis_message"] = "Redis has a high number of idle connections"
    }

    poolUtilization := float64(poolStats.TotalConns-poolStats.IdleConns) / float64(poolSize) * 100
    highPoolUtilizationThreshold := 90.0
    if poolUtilization > highPoolUtilizationThreshold {
        stats["redis_message"] = "Redis connection pool utilization is high"
    }

    return stats
}