Metrics

The Go client provides two levels of metrics: ClusterStats and Extended.

ClusterStats

ClusterStats is an on-demand snapshot of a cluster’s thread and connection usage. To obtain ClusterStats on an active Aerospike Go client instance:

// Takes snapshot of current collected stats.
stats, _ := client.Stats()

ClusterStats (counters)

circuit-breaker-hits: Circuit breaker event counter.
closed-connections: Closed connection counter.
connections-attempts: Number of connection attempts.
connections-error-other: Number of other connection errors.
connections-error-timeout: Number of timed out connection.
connections-failed: Number of failed connections.
connections-idle-dropped: Number of idle connections dropped.
connections-pool-empty: Number of connection empty event.
connections-pool-overflow: Number of connection overflow events.
connections-successful: Number of successful connections.
node-added-count: Number of nodes added.
node-removed-count: Number of nodes removed.
open-connections: Number of opened connections.
partition-map-update: Number of partitions map updates.
tends-failed: Number of tend failed events.
tends-successful: Number of tend successful events.
tends-total: Total number of tend cycles.
transaction-error-count: Number of failed transactions.
transaction-retry-count: Number of retries transactions.
total-nodes: Number of nodes in the cluster.

Extended Metrics

To use extended metrics, you must explicitly notify the client to track latency and command counts for every node. By default, extended metrics are disabled.

To enable:

// Default Metrics policy
// HistogramType:  histogram.Logarithmic,
// LatencyColumns: 24,
// LatencyBase:    2,
client.EnableMetrics(as.DefaultMetricsPolicy())

To disable:

client.DisableMetrics()

The MetricsPolicy fields are:

HistogramType: Histogram type specifies if the histogram should be histogram.Logarithmic(default) or histogram.Linear.
LatencyColumns: LatencyColumns defines the number of elapsed time range buckets in latency histograms.
LatencyBase: Depending on type of the histogram:
- histogram.Logarithmic: base ^ 1, base ^ 2, base ^ 3, … base ^ (columns - 1).
- histogram.Linear: base * 1, base * 2, base * 3, … base * (columns - 1).

`HistogramType:histogram.Logarithmic`, `LatencyColumns:5`, `LatencyBase:2`
2, 4, 8, 16, 32

`HistogramType:histogram.Linear`, `LatencyColumns:5`, `LatencyBase:2`
2, 4, 6, 8, 10

Latency Buckets (histograms)

batch-read-metrics: Metrics for batch read commands.
batch-write-metrics: Metrics for batch write commands.
batch-delete-metrics: Metrics for batch delete commands.
exists-metrics: Metrics for exists commands.
get-header-metrics: Metrics for GetHeader commands.
get-metrics: Metrics for Get commands.
operate-metrics: Metrics for Operate commands.
put-metrics: Metrics for Put commands.
query-metrics: Metrics for Query commands.
scan-metrics: Metrics for Scan commands.
udf-metrics: Metrics for UDF commands.

For each of the histogram metrics we track aggregate values such count, sum, min, and max.

Below is a JSON example of a metrics snapshot. The corresponding metric policy includes the following configuration parameters: HistogramType: histogram.Logarithmic, LatencyColumns: 5, and LatencyBase: 2.

{"127.0.0.1:3000":{"batch-read-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"batch-write-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"circuit-breaker-hits":0,"closed-connections":0,"connections-attempts":2,"connections-error-other":0,"connections-error-timeout":0,"connections-failed":0,"connections-idle-dropped":0,"connections-pool-empty":0,"connections-pool-overflow":0,"connections-successful":2,"delete-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"exists-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"get-header-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"get-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"node-added-count":1,"node-removed-count":0,"open-connections":2,"operate-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"partition-map-updates":1,"put-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"query-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"scan-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"tends-failed":0,"tends-successful":41,"tends-total":41,"transaction-error-count":1,"transaction-retry-count":1,"udf-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0}},"cluster-aggregated-stats":{"batch-read-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"batch-write-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"circuit-breaker-hits":0,"closed-connections":0,"connections-attempts":2,"connections-error-other":0,"connections-error-timeout":0,"connections-failed":0,"connections-idle-dropped":0,"connections-pool-empty":0,"connections-pool-overflow":0,"connections-successful":2,"delete-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"exceeded-max-retries":0,"exceeded-total-timeout":0,"exists-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"get-header-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"get-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"node-added-count":1,"node-removed-count":0,"open-connections":2,"operate-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"partition-map-updates":1,"put-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"query-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"scan-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0},"tends-failed":0,"tends-successful":41,"tends-total":41,"transaction-error-count":1,"transaction-retry-count":1,"udf-metrics":{"buckets":[0,0,0,0,0],"count":0,"max":0,"min":0,"sum":0}},"open-connections":2,"total-nodes":1}