Redis

Integrating Redis Metrics to KubeSense

KubeSense provides comprehensive monitoring for Redis through OpenTelemetry (OTEL) Collector integration, allowing you to collect key Redis metrics and visualize them in the KubeSense platform.

Prerequisites

Before you begin, ensure you have:

A Redis server running version 3.0 or newer
KubeSense deployed with OpenTelemetry (OTEL) Collector with access to the Redis server
- Install the OTEL Collector if not done already
- Ensure the OTEL collector has access to the Redis server for metrics collection

Locating the KubeSense Sensor Service Endpoint

To find the endpoint for the Sensor service (referred to as {KUBESENSOR_HOST}), follow the instructions provided here.

Step 1: Create the Collector Config File

Create a file named redis-metrics-collection-config.yaml with the following content:

receivers:
  redis:
    # The hostname and port of the Redis instance, separated by a colon.
    endpoint: ${env:REDIS_ENDPOINT}
    # The frequency at which to collect metrics from the Redis instance.
    collection_interval: 60s
    # The password used to access the Redis instance; must match the password specified in the requirepass server configuration option.
    password: ${env:REDIS_PASSWORD}
    metrics:
      redis.maxmemory:
        enabled: true
      redis.cmd.latency:
        enabled: true

processors:
  # enriches the data with additional host information
  resourcedetection/system:
    detectors: ["system"]
    system:
      hostname_sources: ["os"]

service:
  pipelines:
    metrics/redis:
      receivers: [redis]
      processors: [resourcedetection/system]
      exporters: [otlphttp/kubesense_redis]

Step 2: Set Environment Variables

Set the following environment variables:

# Redis endpoint reachable from the OTEL collector
export REDIS_ENDPOINT="localhost:6379"

# Password used to access the Redis instance.
# Can be left empty if the redis server is not configured to require a password.
export REDIS_PASSWORD=""

If you want to add custom resource attributes to identify the Redis instance:

processors:
  resource:
    attributes:
    - action: insert
      key: kubesense.cluster
      value: {CLUSTER_NAME}
    - action: insert
      key: kubesense.env_type
      value: (k8s|docker|legacy)

Metrics Pipeline:

pipelines:
  metrics:
    receivers: [redis]
    processors: [resourcedetection/system, resource]
    exporters: [otlphttp/kubesense_redis]

Step 3: Use the Collector Config File

Add the following flag to your collector run command:

--config redis-metrics-collection-config.yaml

Note: The collector can use multiple config files by specifying multiple --config flags.

Redis Metrics Collected

KubeSense automatically collects and enriches the following key Redis metrics:

Metric Name	Category	Description
redis.commands_processed	Performance	Total number of commands processed by the server
redis.commands	Performance	Number of commands processed per second
redis.cmd.latency	Performance	Command execution latency
redis.cpu_time	Performance	System CPU consumed by the Redis server
redis.memory_used	Memory	Total bytes allocated by Redis using its allocator
redis.memory_rss	Memory	Bytes allocated as seen by the operating system
redis.memory_peak	Memory	Peak memory consumed by Redis
redis.memory_lua	Memory	Bytes used by the Lua engine
redis.maxmemory	Memory	Value of the maxmemory configuration directive
redis.memory_fragmentation_ratio	Memory	Ratio between used_memory_rss and used_memory
redis.clients_connected	Connection	Number of client connections (excluding replicas)
redis.clients_blocked	Connection	Number of clients pending on blocking calls
redis.clients_max_input_buffer	Connection	Biggest input buffer among client connections
redis.clients_max_output_buffer	Connection	Longest output list among client connections
redis.connections_received	Connection	Total connections accepted by the server
redis.connections_rejected	Connection	Connections rejected due to maxclients limit
redis.slaves_connected	Connection	Number of connected replicas
redis.net_input	Network	Total bytes read from the network
redis.net_output	Network	Total bytes written to the network
redis.keyspace_hits	Key Space	Successful key lookups in the main dictionary
redis.keyspace_misses	Key Space	Failed key lookups in the main dictionary
redis.db_keys	Key Space	Number of keyspace keys
redis.db_expires	Key Space	Keyspace keys with an expiration
redis.db_avg_ttl	Key Space	Average keyspace key TTL
redis.replication_offset	Replication	Server's current replication offset
redis.replication_backlog_first_byte_offset	Replication	Master offset of the replication backlog buffer
redis.keys_evicted	Replication	Keys evicted due to maxmemory limit
redis.keys_expired	Replication	Total key expiration events
redis.rdb_changes_since_last_save	Replication	Changes since last dump
redis.latest_fork	Replication	Duration of the latest fork operation
redis.uptime	Server	Seconds since Redis server start

Visualization in KubeSense

Once configured, Redis metrics will appear in the KubeSense dashboard, allowing you to:

Monitor Redis performance in real-time
Track memory usage and identify potential bottlenecks
Analyze command patterns and latency
Monitor connection pool health
Set up alerts for key Redis metrics

KubeSense provides seamless Redis monitoring through OpenTelemetry integration, enabling you to track the health and performance of your Redis instances alongside your other infrastructure and applications in a unified observability platform.