{
  "id": "resource-usage",
  "title": "Resource usage metrics",
  "url": "https://redis.io/docs/latest/operate/rs/8.0/references/metrics/resource-usage/",
  "summary": "",
  "content": "\nThe following metrics are available in the Redis Software Cluster Manager UI.\n\n## Connections\n\nNumber of connections to the database.\n\n**Components measured**: Cluster, Node, and Database\n\n## CPU usage\n\nPercent of the node CPU used. \n\n**Components measured**:  Cluster and Node\n\n### Main thread CPU usage \n\nPercent of the CPU used by the main thread.\n\n**Components measured**: Database and Shard\n\n### Fork CPU usage\n\nCPU usage of Redis child forks.\n\n**Components measured**: Database and Shard\n\n### Total CPU usage \n\nPercent usage of the CPU for all nodes.\n\n**Components measured**: Database\n\n## Free disk space\n\nRemaining unused disk space.\n\n**Components measured**:  Cluster and Node\n\n## Memory\n### Used memory \n\nTotal memory used by the database, including RAM, [Flash](https://redis.io/docs/latest/operate/rs/databases/flash) (if enabled), and [replication](https://redis.io/docs/latest/operate/rs/databases/durability-ha/replication) (if enabled).\n\nUsed memory does not include:\n\n1. Fragmentation overhead - The ratio of memory seen by the operating system to memory allocated by Redis\n2. Replication buffers at the primary nodes - Set to 10% of used memory and is between 64 MB and 2048 MB\n3. Memory used by Lua scripts - Does not exceed 1 MB\n4. Copy on Write (COW) operation that can be triggered by:\n    - A full replication process\n    - A database snapshot process\n    - AOF rewrite process\n\nUsed memory is not measured during [shard migration](https://redis.io/docs/latest/operate/rs/databases/configure/replica-ha).\n\nAdditional factors for [Active-Active databases](https://redis.io/docs/latest/operate/rs/databases/active-active):\n\n- By default, the [Active-Active replication backlog](https://redis.io/docs/latest/operate/rs/databases/active-active/manage#replication-backlog), used for synchronization between clusters or regions, is set to 1% of the database size.\n\n- Active-Active databases have a lower threshold for activating the eviction policy, because it requires propagation to all participating clusters. The eviction policy starts to evict keys when one of the Active-Active instances reaches 80% of its memory limit.\n\n**Components measured**: Database and Shard\n\n### Free RAM\n\nAvailable RAM for System use.\n\n**Components measured**:  Cluster and Node\n\n### Memory limit \n\nMemory size limit of the database, enforced on the [used memory](#used-memory).\n\n**Components measured**: Database\n\n### Memory usage \n\nPercent of memory used by Redis out of the [memory limit](#memory-limit).\n\n**Components measured**: Database\n## Traffic\n\n### Incoming traffic \n\nTotal incoming traffic to the database in bytes/sec.\n\nAll incoming traffic is not measured during [shard migration](https://redis.io/docs/latest/operate/rs/databases/configure/replica-ha).\n\n**Components measured**: Cluster, Node and Database\n\n#### Incoming traffic compressed\n\nTotal incoming compressed traffic (in bytes/sec) per [Active-Active](https://redis.io/docs/latest/operate/rs/databases/active-active) replica database. \n\n#### Incoming traffic uncompressed\n\nTotal incoming uncompressed traffic (in bytes/sec) per [Active-Active](https://redis.io/docs/latest/operate/rs/databases/active-active) replica database. \n\n### Outgoing traffic \n\nTotal outgoing traffic from the database in bytes per second.\n\nOutgoing traffic is not measured during [shard migration](https://redis.io/docs/latest/operate/rs/databases/configure/replica-ha).\n\n**Components measured**: Cluster, Node and Database\n\n\n\n\n\n\n\n \n",
  "tags": ["docs","operate","rs"],
  "last_updated": "2026-07-23T12:05:56-05:00"
}