Info

License: Free — part of the universal collection tier.

Redis / Valkey

The redis probe monitors Redis (or Valkey) via the INFO command, reporting memory usage, client connections, command throughput, cache hit/miss ratio, keyspace size, replication state and persistence (RDB/AOF) health.

Quick start

# probes.d/10-redis.yaml — each file under probes.d/ is a YAML array of probes
- name: redis
  type: redis
  params:
    host: 127.0.0.1
    port: 6379

Parameters

Parameter	Must set	Default	Description
host	In practice	127.0.0.1	Server hostname or address
port	In practice	6379	Server port
password	In practice	-	AUTH password, when required. A secret: reference it with ${secret:…}, ${env:…} or ${file:…} rather than writing it in the file
tls	No	false	Use TLS for the connection
tls_cert_file	No	-	Client certificate (PEM) for mutual TLS; needs tls_key_file and tls: true
tls_key_file	No	-	Private key (PEM) matching tls_cert_file
tls_ca_file	No	-	CA bundle (PEM) to verify the server; system trust store by default
timeout	No	5	Connection and command timeout in seconds
interval	No	60	Seconds between collections
instance_name	No	-	Stable identity instead of host:port

Set instance_name to keep the entity identity stable when the address changes.

Metrics

Metric	Unit	Description
senhub.db.up	1	1 when the agent reached the Redis server
redis.uptime	s	Server uptime
redis.clients.connected	{connection}	Connected client count
redis.clients.blocked	{connection}	Clients blocked on a command (BLPOP, etc.)
redis.connections.received	{connection}	Total connections accepted since start
redis.memory.used	By	Memory currently allocated by Redis
redis.memory.peak	By	Peak memory allocation
redis.memory.fragmentation.ratio	1	Memory fragmentation ratio
redis.commands.processed	{command}	Total commands processed since start
redis.keyspace.hits	{hit}	Successful key lookups
redis.keyspace.misses	{miss}	Failed key lookups
redis.expired_keys	{key}	Keys expired since start
redis.evicted_keys	{key}	Keys evicted due to maxmemory policy
redis.replication.lag	s	Replica lag in seconds (replica instances only)
redis.rdb.last_bgsave.duration	s	Duration of the last RDB background save
redis.rdb.last_bgsave.status	{status}	Last RDB save outcome (1 ok, 0 err)
redis.aof.last_bgrewrite.status	{status}	Last AOF rewrite outcome (1 ok, 0 err)
redis.aof.last_write.status	{status}	Last AOF write outcome (1 ok, 0 err)
redis.aof.last_rewrite.duration	s	Duration of the last AOF rewrite, -1 when none has run
redis.replication.backlog_active	{status}	1 when the replication backlog is allocated
redis.replication.backlog_size	By	Configured replication backlog size
redis.replication.backlog_histlen	By	History held in the replication backlog
redis.pubsub.channels	{channel}	Active pub/sub channels
redis.pubsub.patterns	{pattern}	Active pub/sub pattern subscriptions
redis.sentinel.master.status	{status}	Sentinel mode only: one series per monitored master (attribute redis.sentinel.master.name), 1 ok, 0 sdown/odown
redis.sentinel.master.slaves	{replica}	Sentinel mode only: replicas known per master
redis.sentinel.master.sentinels	{sentinel}	Sentinel mode only: sentinels watching each master

Operational notes

• Valkey (the Redis fork) exposes the same INFO interface; configure it identically.
• For Redis Cluster, point the probe at one node; cluster-wide stats are reported per node, not aggregated.
• Metric names align with the community redis_exporter baseline.

Metric reference

Every metric this probe can emit. Metric is the OpenTelemetry name the OTLP, Prometheus and Zabbix outputs derive theirs from. Name is what a Nagios check and the API metrics= filter match. PRTG channel is the label PRTG shows, placeholders filled from the series' tags.

Metric	Name	PRTG channel	Unit	Description
senhub.db.up	senhub.db.up	Redis Up	#	1 if the agent reached the Redis server, 0 otherwise
redis.uptime	redis.uptime	Uptime	s	Seconds since the Redis server started (uptime_in_seconds)
senhub.db.version.info	senhub.db.version.info	Version	#	Redis version banner — value=1, version string carried in attribute db.system.version
redis.clients.connected	redis.clients.connected	Clients Connected	#	Number of client connections (connected_clients)
redis.clients.blocked	redis.clients.blocked	Clients Blocked	#	Clients blocked waiting on a blocking call (blocked_clients)
redis.connections.received	redis.connections.received	Connections Received	#	Cumulative accepted connections (total_connections_received)
redis.connections.rejected	redis.connections.rejected	Connections Rejected	#	Cumulative rejected connections — maxclients cap (rejected_connections)
redis.memory.used	redis.memory.used	Memory Used	B	Allocator-reported used memory (used_memory)
redis.memory.used.rss	redis.memory.used.rss	Memory RSS	B	OS-reported RSS (used_memory_rss) — includes fragmentation overhead
redis.memory.peak	redis.memory.peak	Memory Peak	B	Peak allocator-reported memory (used_memory_peak)
redis.memory.fragmentation.ratio	redis.memory.fragmentation.ratio	Memory Fragmentation Ratio	#	used_memory_rss / used_memory — greater than 1.5 indicates high fragmentation
redis.commands.processed	redis.commands.processed	Commands Processed	#	Cumulative commands processed (total_commands_processed)
redis.net.input	redis.net.input	Network Input	B	Cumulative bytes received from clients (total_net_input_bytes)
redis.net.output	redis.net.output	Network Output	B	Cumulative bytes sent to clients (total_net_output_bytes)
redis.commands	redis.ops.per_sec	Ops/s	#	Instantaneous commands per second (instantaneous_ops_per_sec)
redis.keyspace.hits	redis.keyspace.hits	Keyspace Hits	#	Cumulative successful key lookups (keyspace_hits)
redis.keyspace.misses	redis.keyspace.misses	Keyspace Misses	#	Cumulative failed key lookups (keyspace_misses)
redis.keyspace.hit.ratio	redis.keyspace.hit.ratio	Hit Ratio	%	keyspace_hits / (keyspace_hits + keyspace_misses) — derived gauge, 0 when no traffic
redis.db.keys	redis.db.keys	DB {db} Keys	#	Total keys in logical database (keyspace section: dbN:keys=K)
redis.db.expires	redis.db.expires	DB {db} Expiring Keys	#	Keys with a TTL in logical database (keyspace section: dbN:expires=M)
redis.replication.role	redis.replication.role	Replication Role	#	Instance role: master=1, slave/replica=0, sentinel=-1
redis.replication.offset	redis.replication.offset	Replication Offset	B	Replication offset in bytes (master_repl_offset on master, slave_repl_offset / replica_repl_offset on replica)
redis.replication.slaves.connected	redis.replication.slaves.connected	Replicas Connected	#	Number of replicas currently connected (connected_slaves, master only)
redis.replication.lag	redis.replication.lag	Replication Lag	s	Seconds since last master communication (master_last_io_seconds_ago, replica only)
redis.rdb.changes	redis.rdb.changes	RDB Changes Since Save	#	Writes since last RDB save (rdb_changes_since_last_save)
redis.aof.enabled	redis.aof.enabled	AOF Enabled	#	1 when AOF persistence is enabled (aof_enabled)
redis.rdb.last_bgsave.duration	redis.rdb.last_bgsave.duration	RDB Last BGSave Duration	s	Duration of the last RDB background save in seconds (rdb_last_bgsave_time_sec)
redis.rdb.last_save.age	redis.rdb.last_save.age	RDB Last Save Age	s	Seconds elapsed since the last successful RDB save (now − rdb_last_save_time)
redis.rdb.last_bgsave.status	redis.rdb.last_bgsave.status	RDB Last BGSave Status	#	Outcome of the last RDB background save: 1=ok, 0=err (rdb_last_bgsave_status)
redis.aof.last_bgrewrite.status	redis.aof.last_bgrewrite.status	AOF Last Rewrite Status	#	Outcome of the last AOF background rewrite: 1=ok, 0=err (aof_last_bgrewrite_status)
redis.aof.last_write.status	redis.aof.last_write.status	AOF Last Write Status	#	Outcome of the last write to the AOF: 1=ok, 0=err (aof_last_write_status)
redis.aof.last_rewrite.duration	redis.aof.last_rewrite.duration	AOF Last Rewrite Duration	s	Duration of the last AOF rewrite in seconds, -1 when none has run yet (aof_last_rewrite_time_sec)
redis.latest.fork	redis.latest.fork	Latest Fork Duration	μs	Duration of the latest fork operation in microseconds (latest_fork_usec)
redis.cpu.time	redis.cpu.time	CPU Time ({state})	s	Cumulative CPU time consumed by the Redis server in the given state (used_cpu_sys / used_cpu_user / used_cpu_sys_children / used_cpu_user_children)
redis.memory.lua	redis.memory.lua	Memory Lua	B	Memory used by the Lua scripting engine (used_memory_lua)
redis.clients.max_input_buffer	redis.clients.max_input_buffer	Client Max Input Buffer	B	Largest input buffer across all current client connections (client_recent_max_input_buffer)
redis.clients.max_output_buffer	redis.clients.max_output_buffer	Client Max Output Buffer	B	Largest output buffer across all current client connections (client_recent_max_output_buffer)
redis.db.avg_ttl	redis.db.avg_ttl	DB {db} Average TTL	ms	Average TTL in milliseconds of keys with an expiry in logical database (keyspace section: dbN:avg_ttl=T)
redis.replication.backlog_first_byte_offset	redis.replication.backlog_first_byte_offset	Replication Backlog First Byte Offset	B	Replication backlog first byte offset (repl_backlog_first_byte_offset)
redis.replication.backlog_active	redis.replication.backlog_active	Replication Backlog Active	#	1 when the replication backlog is allocated (repl_backlog_active)
redis.replication.backlog_size	redis.replication.backlog_size	Replication Backlog Size	B	Configured size of the replication backlog (repl_backlog_size)
redis.replication.backlog_histlen	redis.replication.backlog_histlen	Replication Backlog Used	B	Bytes of history currently held in the replication backlog (repl_backlog_histlen)
redis.pubsub.channels	redis.pubsub.channels	Pub/Sub Channels	#	Number of active pub/sub channels (pubsub_channels)
redis.pubsub.patterns	redis.pubsub.patterns	Pub/Sub Patterns	#	Number of active pub/sub pattern subscriptions (pubsub_patterns)
redis.evicted_keys	redis.evicted_keys	Evicted Keys	#	Cumulative keys evicted due to maxmemory policy (evicted_keys)
redis.expired_keys	redis.expired_keys	Expired Keys	#	Cumulative keys expired by the TTL mechanism (expired_keys)
redis.cmd.calls	redis.cmd.calls	Cmd {cmd} Calls	#	Cumulative call count for the given Redis command (INFO commandstats: cmdstat_X:calls=N)
redis.cmd.time	redis.cmd.usec	Cmd {cmd} Usec	μs	Cumulative microseconds spent executing the given Redis command (INFO commandstats: cmdstat_X:usec=N)
redis.cluster.state	redis.cluster.state	Cluster State	#	Cluster health: 1=ok, 0=fail (cluster_state from INFO cluster; emitted only when cluster_enabled=1)
redis.cluster.slots.assigned	redis.cluster.slots.assigned	Cluster Slots Assigned	#	Number of slots assigned to cluster nodes (cluster_slots_assigned)
redis.cluster.slots.ok	redis.cluster.slots.ok	Cluster Slots OK	#	Number of slots not in FAIL or PFAIL state (cluster_slots_ok)
redis.cluster.slots.pfail	redis.cluster.slots.pfail	Cluster Slots PFAIL	#	Number of slots in PFAIL state (cluster_slots_pfail)
redis.cluster.slots.fail	redis.cluster.slots.fail	Cluster Slots FAIL	#	Number of slots in FAIL state — serving requests is blocked until fixed (cluster_slots_fail)
redis.cluster.links.created	redis.cluster.links.created	Cluster Messages Sent	#	Cumulative cluster bus messages sent (cluster_stats_messages_sent)
redis.cluster.links.disconnected	redis.cluster.links.disconnected	Cluster Messages Received	#	Cumulative cluster bus messages received (cluster_stats_messages_received)
redis.sentinel.masters	redis.sentinel.masters	Sentinel Masters	#	Number of Redis masters monitored by this Sentinel (sentinel_masters; emitted only in sentinel mode)
redis.sentinel.slaves	redis.sentinel.slaves	Sentinel Slaves (total)	#	Total number of replicas across all monitored masters (aggregated from per-master lines)
redis.sentinel.ok_slaves	redis.sentinel.ok_slaves	Sentinel Slaves OK	#	Replicas belonging to masters whose status=ok (aggregated from per-master lines)
redis.sentinel.sentinels	redis.sentinel.sentinels	Sentinel Sentinels (total)	#	Total number of Sentinel peers across all monitored masters (aggregated from per-master lines)
redis.sentinel.ok_sentinels	redis.sentinel.ok_sentinels	Sentinel Sentinels OK	#	Sentinel peers belonging to masters whose status=ok (aggregated from per-master lines)
redis.sentinel.scripts_queue_length	redis.sentinel.scripts_queue_length	Sentinel Scripts Queue	#	Number of scripts in the Sentinel notification-scripts queue (sentinel_running_scripts)
redis.sentinel.master.status	redis.sentinel.master.status	Sentinel Master {master} Status	#	Status of a monitored master: 1=ok, 0=odown/sdown or any other state (INFO sentinel masterN status; one series per master, attribute redis.sentinel.master.name)
redis.sentinel.master.slaves	redis.sentinel.master.slaves	Sentinel Master {master} Replicas	#	Replicas known for a monitored master (INFO sentinel masterN slaves)
redis.sentinel.master.sentinels	redis.sentinel.master.sentinels	Sentinel Master {master} Sentinels	#	Sentinels monitoring a master, this one included (INFO sentinel masterN sentinels)
redis.tracking.clients	redis.tracking.clients	Tracking Clients	#	Number of clients using RESP3 client-side caching tracking (tracking_clients; 0 when absent)
redis.tracking.keys	redis.tracking.keys	Tracking Keys	#	Number of keys in the client-side tracking invalidation table (tracking_table_used_keys; 0 when absent)