Redshift Metrics

For each Redshift cluster, two types of elements are collected:

  • RedshiftCluster: Contains cluster-specific metrics as well as metrics that are averages across all nodes.

  • RedshiftNode: Contains node-specific metrics. There will be one element per Redshift node.

The table below denotes which metrics are cluster- or node-based (or both).

Collected

Fully Qualified Name (FQN)

Cluster

Node

AWS Metric

Statistic

Units

Max

BASE

CORR

UTIL

aws.redshift.cpuutilization

yes

yes

CPUUtilization

average

percent

100

yes

yes

yes

aws.redshift.databaseconnections

yes

no

DatabaseConnections

average

count

none

yes

no

no

aws.redshift.healthstatus

yes

no

HealthStatus

average

1

no

no

no

aws.redshift.maintenancemode

yes

no

MaintenanceMode

average

1

no

no

no

aws.redshift.networkreceivethroughput

yes

yes

NetworkReceiveThroughput

average

Bps

none

yes

yes

no

aws.redshift.networktransmitthroughput

yes

yes

NetworkTransmitThroughput

average

Bps

none

yes

yes

no

aws.redshift.percetagediskspaceused

yes

yes

PercentageDiskSpaceUsed

average

percent

100

yes

no

yes

aws.redshift.readiops

yes

yes

ReadIOPS

average

iops

none

yes

yes

no

aws.redshift.readlatency

yes

yes

ReadLatency

average

seconds

none

yes

yes

no

aws.redshift.readthroughput

yes

yes

ReadThroughput

average

Bps

none

yes

yes

no

aws.redshift.writeiops

yes

yes

WriteIOPS

average

iops

none

yes

yes

no

aws.redshift.writelatency

yes

yes

WriteLatency

average

seconds

none

yes

yes

no

aws.redshift.writethroughput

yes

yes

WriteThroughput

average

Bps

none

yes

yes

no

Computed

Fully Qualified Name (FQN)

Description

Units

BASE

netuitive.aws.redshift.totalthroughput

This metric represents the total throughput, obtained by adding the read and write throughputs. Computation:(ReadThroughput + WriteThroughput)

Bps

yes