9.7.1.2.1 Real-Time Insight Attributes

Commencing with Oracle Exadata System Software release 22.1.0, you can use the Real-Time Insight feature to enable real-time monitoring of your Exadata systems.

Fine-Grained Metric Collection

The metricFGCollIntvlInSec attribute controls fine-grained metric collection.

  • To enable fine-grained metric collection, you must set the collection interval to a value between 1 and 60 seconds.

    For example:

    DBMCLI> ALTER DBSERVER metricFGCollIntvlInSec=1

    The metricFGCollIntvlInSec setting is related to the automatic upload frequency specified in the metricStreamIntvlInSec attribute. When automatic metric upload and fine-grained collection are both enabled (metricStreamIntvlInSec>0 and metricFGCollIntvlInSec>0), metricStreamIntvlInSec must be between 5 and 30 times metricFGCollIntvlInSec. For example, if metricStreamIntvlInSec=60, then metricFGCollIntvlInSec must be between 2 and 12.

  • To disable fine-grained metric collection on a server, set metricFGCollIntvlInSec=0.

    For example:

    DBMCLI> ALTER DBSERVER metricFGCollIntvlInSec=0

    Fine-grained metric collection can be disabled only when automatic metric upload is disabled (metricStreamIntvlInSec=0) or the automatic upload frequency is between 5 and 30 minutes (metricStreamIntvlInSec is between 300 and 1800).

Automatic Metric Upload

The metricStreamIntvlInSec attribute sets the upload interval (in seconds) for automatic uploads to the metric streaming endpoints specified by the metricStreamEndPoint attribute.

  • To enable automatic metric uploads, set the metricStreamIntvlInSec attribute to a non-zero value.

    For example:

    DBMCLI> ALTER DBSERVER metricStreamIntvlInSec=25

    The metricStreamIntvlInSec setting is related to the fine-grained collection frequency specified in the metricFGCollIntvlInSec attribute:

    • When automatic metric upload and fine-grained collection are both enabled (metricStreamIntvlInSec>0 and metricFGCollIntvlInSec>0), metricStreamIntvlInSec must be between 5 and 30 times metricFGCollIntvlInSec. For example, if metricFGCollIntvlInSec is set to 5, then metricStreamIntvlInSec must be between 25 and 150.

    • When automatic metric upload is enabled and fine-grained collection is disabled (metricStreamIntvlInSec>0 and metricFGCollIntvlInSec=0), the automatic upload frequency must be between 5 and 30 minutes (metricStreamIntvlInSec must be between 300 and 1800).

  • To disable automatic metric uploads, set metricStreamIntvlInSec=0.

    For example:

    DBMCLI> ALTER DBSERVER metricStreamIntvlInSec=0

Metric Upload Endpoints

The metricStreamEndPoint attribute specifies one or more collection endpoints that automatically receive the metric stream. You can set metricStreamEndPoint as follows: 

metricStreamEndPoint[+]=((host="endpoint-URL"[,type="stream-format"][,token="authentication-token"]
                            [,{httpProxy|httpsProxy}="proxy-server"][,certFileName="certificate-file-path"])
                          [,(host="endpoint-URL"[,type="stream-format"][,token="authentication-token"]
                            [,{httpProxy|httpsProxy}="proxy-server"][,certFileName="certificate-file-path"])]...)

In the metricStreamEndPoint definition:

  • host: Specifies the URL for the collection endpoint. The URL can use HTTP or HTTPS.

  • type: Optionally specifies the format of the stream. Supported values are:

    • json: Provides the stream in a JSON format

    • plaintext: Provides the stream in a plain text format

    The default value is json.

  • token: Optionally specifies the authentication token for the collection endpoint. Consult the metric collection platform for details about generating the token.

  • httpProxy or httpsProxy: Optionally specifies a proxy server to facilitate network connectivity to the collection endpoint. A proxy server is required if a firewall resides between the Exadata system and the collection endpoint.

  • certFileName: Optionally specifies the local path to a certificate file that enables metric streaming to an HTTPS receiver that uses a self-signed certificate or a certificate issued by a private certificate authority (CA).

    When configured, Management Server (MS) validates the certificate and copies it to its protected configuration directory. MS does not use the original certificate file after copying it. For metric uploads, MS creates an endpoint-specific TLS trust context from the copied certificate. This allows MS to authenticate the HTTPS receiver without requiring the certificate to be added to the default CA trust store.

    This option applies only to HTTPS endpoints. The certificate file must contain a valid X.509 server certificate or certificate chain. PEM encoding is supported. For an endpoint specified by an IP address, the certificate's Subject Alternative Name (SAN) should include that IP address.

    This option is available in Oracle Exadata System Software release 25.2.0 and later.

You can use the optional += operator to add collection endpoints to an existing metricStreamEndPoint definition. Otherwise, the = operator overwrites the previous attribute value.

Metric Tags

The metricStreamTags attribute defines a set of metric tags, which are included in every metric observation generated by the server. These tags can help you to organize and group observations generated by numerous Exadata servers.

You can set the metricStreamTags attribute to a valid JSON string containing tag and value pairs as follows: 

metricStreamTags='{"tag1":"value1"[,"tag2":"value2"]...}'

For example:

DBMCLI> ALTER DBSERVER metricStreamTags='{"application":"personnel","department":"HR"}'

Parent topic: ALTER DBSERVER