AWS Redshift¶
The core performance Metrics of AWS Redshift include query performance, disk space usage, CPU utilization, database connections, and disk I/O operations. These are key Metrics for evaluating and optimizing the performance of a data warehouse.
Configuration¶
Install Func¶
It is recommended to enable TrueWatch integration - extension - DataFlux Func (Automata): all prerequisites will be automatically installed. Please continue with script installation.
If you deploy Func by yourself, refer to Self-deploy Func
Installation Script¶
Note: Please prepare the required Amazon AK in advance (for simplicity, you can directly grant global read-only permission
ReadOnlyAccess
).
To synchronize Redshift monitoring data, we install the corresponding collection script: "TrueWatch Integration (AWS-Redshift Collection)" (ID: truewatch_aws_redshift
)
After clicking [Install], enter the corresponding parameters: Amazon AK, Amazon account name.
Click [Deploy Startup Script], and the system will automatically create a Startup
script set and automatically configure the corresponding startup script.
In addition, you can see the corresponding automatic trigger configuration in [Manage / Automatic Trigger Configuration]. Click [Execute] to run it immediately without waiting for the scheduled time. After a while, you can view the execution task records and corresponding logs.
We collect some configurations by default, for details, see the Metrics column Customize Cloud Object Metrics Configuration
Verification¶
- In [Manage / Automatic Trigger Configuration], confirm whether the corresponding task has the corresponding automatic trigger configuration, and at the same time, you can check the corresponding task records and logs for any abnormalities.
- On the TrueWatch platform, under [Infrastructure / Custom], check if there is asset information.
- On the TrueWatch platform, under [Metrics], check if there is corresponding monitoring data.
Metrics¶
Configure Amazon CloudWatch, the default Measurement set is as follows. You can collect more Metrics through configuration Amazon CloudWatch Metrics Details
Redshift Metrics¶
Metric Name | Description | Unit | Dimensions |
---|---|---|---|
CPUUtilization | Percentage of CPU usage. For clusters, this metric represents the sum of CPU usage values for all nodes (leader node and compute nodes). | % | ClusterIdentifier |
HealthStatus | Cluster health check | 1: Normal or 0: Abnormal | ClusterIdentifier |
MaintenanceMode | Whether the cluster is in maintenance mode | 1: ON or 0: OFF | ClusterIdentifier |
PercentageDiskSpaceUsed | Percentage of used disk space | % | ClusterIdentifier |
DatabaseConnections | Number of database connections in the cluster | count | ClusterIdentifier |
CommitQueueLength | Number of transactions waiting to commit at a given point in time | count | ClusterIdentifier |
ConcurrencyScalingActiveClusters | The number of concurrent scaling clusters actively processing queries at any given time | count | ClusterIdentifier |
NetworkReceiveThroughput | The rate at which nodes or clusters receive data | byte/s | ClusterIdentifier |
NetworkTransmitThroughput | The rate at which nodes or clusters write data | byte/s | ClusterIdentifier |
MaxConfiguredConcurrencyScalingClusters | Maximum number of concurrent scaling clusters configured from parameter groups | count | ClusterIdentifier |
NumExceededSchemaQuotas | Number of schemas exceeding quotas | count | ClusterIdentifier |
ReadIOPS | Average number of disk read operations per second | count/s | ClusterIdentifier |
ReadLatency | Average time required for disk read I/O operations | seconds | ClusterIdentifier |
ReadThroughput | Average number of bytes read from disk per second | byte | ClusterIdentifier |
TotalTableCount | Number of user tables open at a specific point in time | count | ClusterIdentifier |
WriteIOPS | Average number of disk write operations per second | count/s | ClusterIdentifier |
WriteLatency | Average time required for disk write I/O operations | seconds | ClusterIdentifier |
WriteThroughput | Average number of bytes written to disk per second | byte | ClusterIdentifier |
Objects¶
The structure of collected Amazon AWS Redshift object data can be viewed under [Infrastructure - Custom]
{
"measurement": "aws_redshift",
"tags": {
"ClusterAvailabilityStatus": "Modifying",
"ClusterIdentifier": "hn-test",
"ClusterStatus": "creating",
"ClusterSubnetGroupName": "default",
"ClusterVersion": "1.0",
"DBName": "dev",
"MasterUsername": "awsuser",
"NodeType": "dc2.large",
"PreferredMaintenanceWindow": "sat:19:30-sat:20:00",
"RegionId": "cn-northwest-1",
"VpcId": "vpc-b1ca3ff0fa4d",
"name": "hn-test"
},
"fields": {
"AllowVersionUpgrade": true,
"AutomatedSnapshotRetentionPeriod": 1,
"ClusterNodes": "[]",
"ClusterParameterGroups": "[{\"ParameterApplyStatus\": \"in-sync\", \"ParameterGroupName\": \"default.redshift-1.0\"}]",
"ClusterSecurityGroups": "[]",
"Encrypted": false,
"Endpoint": "{}",
"ManualSnapshotRetentionPeriod": -1,
"NumberOfNodes": 1,
"PendingModifiedValues": "{\"MasterUserPassword\": \"****\"}",
"PubliclyAccessible": false,
"VpcSecurityGroups": "[{\"Status\": \"active\", \"VpcSecurityGroupId\": \"sg-467a\"}]",
"message": "{JSON instance data}"
}
}
Note: The fields in
tags
andfields
may change with subsequent updates.Tip 1: The value of
tags.name
is the instance ID, used as a unique identifier.Tip 2:
fields.message
,fields.network_interfaces
,fields.blockdevicemappings
are JSON serialized strings.