Cluster Connection
The agent requires a connection to the cluster in order to gather information about the cluster and data structure. It is very important that the values provided here exactly match the details specified in the Cluster Properties of the AWS console for this Redshift cluster.
- Cluster Name – The name of the Redshift cluster. This is a unique key that identifies the cluster.
- Region – The EC2 Availability Zone that the cluster was created in. Described as Zone in the Cluster Properties.
- IP or Hostname – The Endpoint address for the cluster minus the port number.
- Port – The port number included in the Endpoint address.
- Database – Name of a database to connect to. All databases will be monitored, but data gathered from system tables and views available to the entire cluster will be through this database.
- Username – Redshift database user that must have at a minimum SELECT privilege on all system tables and views.
- Access Key ID – Access Key ID of the IAM user created during pre-configuration or existing user with necessary attached policies.
- Secret Access Key – Secret access key for the user described above.
- SSL Mode – Enforced security for making a connection to the Redshift cluster. If require_ssl is set to true for the cluster parameter group, this must be set to at least Enable.
Collection Intervals
The Collection Interval fields in the agent properties are used to set the sample frequencies. You can turn off a collection by setting the interval to 0. The defaults are set based on the type of data being collected for relevancy. Data gathered from CloudWatch will always be collected at 5-minute intervals.
- DB List Refresh – Checks existing databases in order to remove collection tasks for dropped databases or schedule new collections for newly created databases.
- Availability – Checks ability of the agent to connect to and request data from the cluster.
- Cluster – Retrieves cluster-level data from the database.
- Tables – Retrieves table structure and statistics from each database.
- Configuration – Retrieves configuration values from pg_settings.
DB Overrides
The DB override properties allow you finer-grained control over database collections. The default list is populated with system databases that have restricted use and will cause the agent to generate errors if access attempts are made. If a database is not specified here, it will use the parameters set in the Collection Intervals section.
- Database Name – Name of the database where collections are being overridden.
- Ignore Tables – If set to true, table information for this database will not be collected.
- Interval – This collection interval will override the default Tables setting in the Collection Intervals section for this database.
Statement Tracking
- Enable Statement Tracking – Enables or disables use of statement tracking.
- # of Top Statements – The maximum number of statements for the agent to collect during each sample period.
- Order By – The statements are sorted in order to gather the # of top statements specified in the previous property. If all statements are being gathered, this field is unimportant. Otherwise, several self-explanatory options are available based on existing or calculated column data in the view.
- Enable S3 Query Tracking – Enables statement tracking specifically for S3 queries.
- # of Top S3 Queries – The maximum number of S3 queries for the agent to collect during each sample period.
Options
FMS Time Offset – This option allows conversion of timestamp information from various collections. Timestamp information from Redshift is generally in UTC time. The default value of SYSTEM will automatically convert timestamps based on the offset of the machine that hosts the agent from UTC.