Metrics for Redis
The AWS/ElastiCache namespace includes the following Redis metrics.
With the exception of ReplicationLag and EngineCPUUtilization, these metrics are derived from the Redis
info command. Each metric is calculated at the cache node level.
For complete documentation of the Redis info command, see
http://redis.io/commands/info
See Also
| Metric | Description | Unit |
|---|---|---|
ActiveDefragHits |
The number of value reallocations per minute performed
by the active defragmentation process. This is derived from active_defrag_hits statistic at Redis INFO |
Number |
AuthenticationFailures |
The total number of failed attempts to authenticate to Redis using the AUTH command.
You can find more information about individual authentication failures
using the ACL LOG |
Count |
BytesUsedForCache |
The total number of bytes allocated by Redis for all purposes, including the dataset, buffers, and so on. | Bytes |
Dimension: Tier=Memory for Redis clusters using Data tiering: The total number
of bytes used for cache by memory. This is the value of used_memory statistic at Redis
INFO |
Bytes | |
Dimension: Tier=SSD for Redis clusters using Data tiering: The total number of bytes used for cache by SSD. |
Bytes | |
BytesReadFromDisk |
The total number of bytes read from disk per minute. Supported only for clusters using Data tiering. | Bytes |
BytesWrittenToDisk |
The total number of bytes written to disk per minute. Supported only for clusters using Data tiering. | Bytes |
CacheHits |
The number of successful read-only key lookups in the main dictionary. This is derived from keyspace_hits statistic at Redis INFO |
Count |
CacheMisses |
The number of unsuccessful read-only key lookups in the main dictionary. This is derived from keyspace_misses statistic at Redis INFO |
Count |
CommandAuthorizationFailures |
The total number of failed attempts by users to run commands they don’t have permission
to call. You can find more information about individual authentication
failures using the ACL
LOG |
Count |
CacheHitRate |
Indicates the usage efficiency of the Redis instance. If the cache ratio is lower than
about 0.8, it means that a significant amount of keys are evicted,
expired, or don't exist. This is calculated using
cache_hits and cache_misses statistics in
the following way: cache_hits /(cache_hits +
cache_misses). |
Percent |
CurrConnections |
The number of client connections, excluding connections from read
replicas. ElastiCache uses two to four of the connections to monitor the
cluster in each case. This is derived from the connected_clients statistic at Redis INFO |
Count |
CurrItems |
The number of items in the cache. This is derived from the Redis
keyspace statistic, summing all of the keys in the entire
keyspace. |
Count |
Dimension: Tier=Memory for clusters using Data tiering. The number of items in memory. |
Count | |
Dimension: Tier=SSD (solid state drives) for clusters using Data tiering. The number of items in SSD. |
Count | |
CurrVolatileItems |
Total number of keys in all databases that have a ttl set. This is derived from the Redis
expires statistic, summing all of the keys with a ttl set in the entire
keyspace. |
Count |
DatabaseMemoryUsagePercentage |
Percentage of the memory available for the cluster that is in use. This is calculated using used_memory/maxmemory from Redis INFO |
Percent |
DatabaseMemoryUsageCountedForEvictPercentage |
Percentage of the memory available for the cluster that is in use excluding memory used for overhead and COB. This is calculated using used_memory-mem_not_counted_for_evict/maxmemory from Redis INFO |
Percent |
DB0AverageTTL |
Exposes avg_ttl of DBO from the keyspace statistic of Redis INFO |
Milliseconds |
EngineCPUUtilization |
Provides CPU utilization of the Redis engine thread. Because Redis is single-threaded, you can
use this metric to analyze the load of the Redis process itself. The
On an ElastiCache host, background processes monitor the host to provide a managed database
experience. These background processes can take up a significant
portion of the CPU workload. This is not significant on larger
hosts with more than two vCPUs. But it can affect smaller hosts
with 2vCPUs or fewer. If you only monitor the
|
Percent |
Evictions |
The number of keys that have been evicted due to the
maxmemory limit. This is derived from the evicted_keys statistic at Redis INFO |
Count |
GlobalDatastoreReplicationLag |
This is the lag between the secondary Region's primary node and the primary Region's primary node. For cluster mode enabled Redis, the lag indicates the maximum delay among the shards. | Seconds |
IsMaster |
Indicates whether the node is the primary node of current shard/cluster. The metric can be either 0 (not primary) or 1 (primary). | Count |
KeyAuthorizationFailures |
The total number of failed attempts by users to access keys they don’t have permission
to access. You can find more information about individual authentication
failures using the ACL
LOG |
Count |
KeysTracked |
The number of keys being tracked by Redis key tracking as a percentage of
tracking-table-max-keys. Key tracking is used to aid
client-side caching and notifies clients when keys are modified.
|
Count |
MemoryFragmentationRatio |
Indicates the efficiency in the allocation of memory of the Redis engine. Certain
thresholds signify different behaviors. The recommended value is to have
fragmentation above 1.0. This is calculated from the
mem_fragmentation_ratio statistic of Redis INFO |
Number |
NewConnections |
The total number of connections that have been accepted by the server
during this period. This is derived from the total_connections_received statistic at Redis INFO |
Count |
NumItemsReadFromDisk |
The total number of items retrieved from disk per minute. Supported only for clusters using Data tiering. | Count |
NumItemsWrittenToDisk |
The total number of items written to disk per minute. Supported only for clusters using Data tiering. | Count |
MasterLinkHealthStatus |
This status has two values: 0 or 1. The value 0 indicates that data in the ElastiCache primary node is not in sync with Redis on EC2. The value of 1 indicates that the data is in sync. To complete the migration, use the CompleteMigration API operation. | Boolean |
Reclaimed |
The total number of key expiration events. This is derived from the expired_keys statistic at Redis INFO |
Count |
ReplicationBytes |
For nodes in a replicated configuration, ReplicationBytes reports the
number of bytes that the primary is sending to all of its replicas. This
metric is representative of the write load on the replication group.
This is derived from the master_repl_offset statistic at
Redis INFO |
Bytes |
ReplicationLag |
This metric is only applicable for a node running as a read replica. It represents how far behind, in seconds, the replica is in applying changes from the primary node. For Redis engine version 5.0.6 onwards, the lag can be measured in milliseconds. | Seconds |
SaveInProgress |
This binary metric returns 1 whenever a background save (forked or forkless) is in
progress, and 0 otherwise. A background save process is typically used
during snapshots and syncs. These operations can cause degraded
performance. Using the SaveInProgress metric, you can
diagnose whether degraded performance was caused by a background save
process. This is derived from the rdb_bgsave_in_progress
statistic at Redis
INFO |
Count |
EngineCPUUtilization availability
AWS Regions listed following are available on all supported node types.
| Region | Region name |
|---|---|
us-east-2 |
US East (Ohio) |
us-east-1 |
US East (N. Virginia) |
us-west-1 |
US West (N. California) |
us-west-2 |
US West (Oregon) |
ap-northeast-1 |
Asia Pacific (Tokyo) |
ap-northeast-2 |
Asia Pacific (Seoul) |
ap-northeast-3 |
Asia Pacific (Osaka) |
ap-east-1 |
Asia Pacific (Hong Kong) |
ap-south-1 |
Asia Pacific (Mumbai) |
ap-southeast-1 |
Asia Pacific (Singapore) |
ap-southeast-2 |
Asia Pacific (Sydney) |
ap-southeast-3 |
Asia Pacific (Jakarta) |
ca-central-1 |
Canada (Central) |
cn-north-1 |
China (Beijing) |
cn-northwest-2 |
China (Ningxia) |
me-south-1 |
Middle East (Bahrain) |
eu-central-1 |
Europe (Frankfurt) |
eu-west-1 |
Europe (Ireland) |
eu-west-2 |
Europe (London) |
eu-west-3 |
EU (Paris) |
eu-south-1 |
Europe (Milan) |
af-south-1 |
Africa (Cape Town) |
eu-north-1 |
Europe (Stockholm) |
sa-east-1 |
South America (São Paulo) |
us-gov-west-1 |
AWS GovCloud (US-West) |
us-gov-east-1 |
AWS GovCloud (US-East) |
The following are aggregations of certain kinds of commands, derived from info
commandstats. The commandstats section provides statistics based on the command type, including the number of calls, the total CPU time consumed by these commands, and the average CPU consumed per command execution.
For each command type, the following line is added:
cmdstat_XXX: calls=XXX,usec=XXX,usec_per_call=XXX.
The latency metrics listed following are calculated using commandstats statistic from Redis INFOdelta(usec)/delta(calls). delta is calculated as the diff within one minute.
Latency is defined as CPU time taken by ElastiCache to process the command. Note that for clusters using data tiering, the time taken to fetch items from SSD is not included in these measurements.
For a full list of available commands, see redis commands
| Metric | Description | Unit |
|---|---|---|
ClusterBasedCmds |
The total number of commands that are cluster-based. This is derived from the Redis commandstats statistic by
summing all of the commands that act upon a cluster (cluster slot, cluster info, and so on). |
Count |
ClusterBasedCmdsLatency |
Latency of cluster-based commands. | Microseconds |
EvalBasedCmds |
The total number of commands for eval-based commands. This is derived from the Redis
commandstats statistic by summing eval,
evalsha. |
Count |
EvalBasedCmdsLatency |
Latency of eval-based commands. | Microseconds |
GeoSpatialBasedCmds |
The total number of commands for geospatial-based commands. This is derived from the
Redis commandstats statistic. It's derived by summing all
of the geo type of commands:
geoadd, geodist, geohash,
geopos, georadius, and
georadiusbymember. |
Count |
GeoSpatialBasedCmdsLatency |
Latency of geospatial-based commands. | Microseconds |
GetTypeCmds |
The total number of read-only type commands. This is derived from
the Redis commandstats statistic by summing all of the
read-only type commands (get,
hget, scard,
lrange, and so on.) |
Count |
GetTypeCmdsLatency |
Latency of read commands. | Microseconds |
HashBasedCmds |
The total number of commands that are hash-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more hashes (hget,
hkeys, hvals,
hdel, and so on). |
Count |
HashBasedCmdsLatency |
Latency of hash-based commands. | Microseconds |
HyperLogLogBasedCmds |
The total number of HyperLogLog-based commands. This is derived from the
Redis commandstats statistic by summing all of the
pf type of commands (pfadd,
pfcount, pfmerge, and so
on.). |
Count |
HyperLogLogBasedCmdsLatency |
Latency of HyperLogLog-based commands. | Microseconds |
JsonBasedCmds |
The total number of commands that are JSON-based. This is derived from the Redis commandstats statistic by summing all of the commands that act upon one or more JSON document objects. |
Count |
JsonBasedCmdsLatency |
Exposes the aggregate latency (server side CPU time) calculated as Delta[Usec]/Delta[Calls] of all commands that act upon one or more JSON document objects. |
Microseconds |
KeyBasedCmds |
The total number of commands that are key-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more keys across multiple data structures
(del, expire,
rename, and so on.). |
Count |
KeyBasedCmdsLatency |
Latency of key-based commands. | Microseconds |
ListBasedCmds |
The total number of commands that are list-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more lists (lindex,
lrange, lpush,
ltrim, and so on). |
Count |
ListBasedCmdsLatency |
Latency of list-based commands. | Microseconds |
PubSubBasedCmds |
The total number of commands for pub/sub functionality.
This is derived from the Redis commandstatsstatistics by summing all of the commands used for pub/sub functionality:
psubscribe, publish, pubsub, punsubscribe, subscribe, and unsubscribe. |
Count |
PubSubBasedCmdsLatency |
Latency of pub/sub-based commands. | Microseconds |
SetBasedCmds |
The total number of commands that are set-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more sets (scard,
sdiff, sadd,
sunion, and so on). |
Count |
SetBasedCmdsLatency |
Latency of set-based commands. | Microseconds |
SetTypeCmds |
The total number of write types of commands. This is derived from
the Redis commandstats statistic by summing all of the
mutative types of commands that operate on data
(set, hset,
sadd, lpop, and so
on.) |
Count |
SetTypeCmdsLatency |
Latency of write commands. | Microseconds |
SortedSetBasedCmds |
The total number of commands that are sorted set-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more sorted sets (zcount,
zrange, zrank,
zadd, and so on). |
Count |
SortedSetBasedCmdsLatency |
Latency of sorted-based commands. | Microseconds |
StringBasedCmds |
The total number of commands that are string-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more strings (strlen,
setex, setrange, and so
on). |
Count |
StringBasedCmdsLatency |
Latency of string-based commands. | Microseconds |
StreamBasedCmds |
The total number of commands that are stream-based. This is derived from the Redis
commandstats statistic by summing all of the commands
that act upon one or more streams data types (xrange,
xlen, xadd,
xdel, and so on). |
Count |
StreamBasedCmdsLatency |
Latency of stream-based commands. | Microseconds |
