このページは日本語には対応しておりません。随時翻訳に取り組んでいます。
翻訳に関してご質問やご意見ございましたら、お気軽にご連絡ください

Overview

This guide describes setting up the Oracle integration on versions of the Datadog Agent lower than 7.50.1. For more information on the Oracle integration, including setting it up on newer Agent versions, see the Oracle integration documentation.

Setup

Installation

Prerequisite

To use the Oracle integration you can either use the native client (no additional install steps required), or the Oracle Instant Client.

Oracle Instant Client
  1. Follow the Oracle Instant Client installation for Linux.

  2. Verify the following:

    • Both the Instant Client Basic and SDK packages are installed. Find them on Oracle’s download page.

      After the Instant Client libraries are installed, ensure the runtime linker can find the libraries. For example, using ldconfig:

      # Put the library location in an ld configuration file.
      
      sudo sh -c "echo /usr/lib/oracle/12.2/client64/lib > \
          /etc/ld.so.conf.d/oracle-instantclient.conf"
      
      # Update the bindings.
      
      sudo ldconfig
      
    • Both packages are decompressed into a single directory that is available to all users on the given machine (for example, /opt/oracle):

      mkdir -p /opt/oracle/ && cd /opt/oracle/
      unzip /opt/oracle/instantclient-basic-linux.x64-12.1.0.2.0.zip
      unzip /opt/oracle/instantclient-sdk-linux.x64-12.1.0.2.0.zip
      
  1. Follow the Oracle Windows installation guide to configure your Oracle Instant Client.

  2. Verify the following:

    • The Microsoft Visual Studio 2017 Redistributable or the appropriate version is installed for the Oracle Instant Client.

    • Both the Instant Client Basic and SDK packages from Oracle’s [download page][18] are installed.

    • Both packages are extracted into a single directory that is available to all users on the given machine (for example, C:\oracle).

JDBC driver

NOTE: This method only works on Linux.

Java 8 or higher is required on your system for JPype, one of the libraries used by the Agent when using JDBC driver.

Once it is installed, complete the following steps:

  1. Download the JDBC Driver JAR file.
  2. Add the path to the downloaded file in your $CLASSPATH or the check configuration file under jdbc_driver_path (see the sample oracle.yaml).

Datadog user creation

Create a read-only datadog user with proper access to your Oracle Database Server. Connect to your Oracle database with an administrative user, such as SYSDBA or SYSOPER, and run:

-- Enable Oracle Script.
ALTER SESSION SET "_ORACLE_SCRIPT"=true;

-- Create the datadog user. Replace the password placeholder with a secure password.
CREATE USER datadog IDENTIFIED BY <PASSWORD>;

-- Grant access to the datadog user.
GRANT CONNECT TO datadog;
GRANT SELECT ON GV_$PROCESS TO datadog;
GRANT SELECT ON gv_$sysmetric TO datadog;
GRANT SELECT ON sys.dba_data_files TO datadog;
GRANT SELECT ON sys.dba_tablespaces TO datadog;
GRANT SELECT ON sys.dba_tablespace_usage_metrics TO datadog;

Note: If you’re using Oracle 11g, there’s no need to run the following line:

ALTER SESSION SET "_ORACLE_SCRIPT"=true;
Oracle 12c or 19c

Log in to the root database as an Administrator to create a datadog user and grant permissions:

alter session set container = cdb$root;
CREATE USER c##datadog IDENTIFIED BY password CONTAINER=ALL;
GRANT CREATE SESSION TO c##datadog CONTAINER=ALL;
Grant select any dictionary to c##datadog container=all;
GRANT SELECT ON GV_$PROCESS TO c##datadog CONTAINER=ALL;
GRANT SELECT ON gv_$sysmetric TO c##datadog CONTAINER=ALL;

Configuration

To configure this check for an Agent running on a host:

  1. Edit the oracle.d/conf.yaml file, in the conf.d/ folder at the root of your Agent’s configuration directory. Update the server and port to set the masters to monitor. See the sample oracle.d/conf.yaml for all available configuration options.

    init_config:
    
    instances:
       ## @param server - string - required
       ## The IP address or hostname of the Oracle Database Server.
       #
       - server: localhost:1521
    
         ## @param service_name - string - required
         ## The Oracle Database service name. To view the services available on your server,
         ## run the following query: `SELECT value FROM v$parameter WHERE name='service_names'`
         #
         service_name: <SERVICE_NAME>
    
         ## @param username - string - required
         ## The username for the Datadog user account.
         #
         username: <USERNAME>
    
         ## @param password - string - required
         ## The password for the Datadog user account.
         #
         password: <PASSWORD>
    
  2. Restart the Agent.

Only custom queries

To skip default metric checks for an instance and only run custom queries with an existing metrics-gathering user, insert the tag only_custom_queries with a value of true. This allows a configured instance of the Oracle integration to prevent the system, process, and tablespace metrics from running, and allows custom queries to be run without having the permissions described in the Datadog user creation section. If this configuration entry is omitted, the user you specify must have those table permissions to run a custom query.

init_config:

instances:
  ## @param server - string - required
  ## The IP address or hostname of the Oracle Database Server.
  #
  - server: localhost:1521

    ## @param service_name - string - required
    ## The Oracle Database service name. To view the services available on your server,
    ## run the following query:
    ## `SELECT value FROM v$parameter WHERE name='service_names'`
    #
    service_name: "<SERVICE_NAME>"

    ## @param username - string - required
    ## The username for the user account.
    #
    username: <USER>

    ## @param password - string - required
    ## The password for the user account.
    #
    password: "<PASSWORD>"

    ## @param only_custom_queries - string - optional
    ## Set this parameter to any value if you want to only run custom
    ## queries for this instance.
    #
    only_custom_queries: true

Connect to Oracle through TCPS

  1. To connect to Oracle through TCPS (TCP with SSL), uncomment the protocol configuration option and select TCPS. Update the server option to set the TCPS server to monitor.

    init_config:
    
    instances:
      ## @param server - string - required
      ## The IP address or hostname of the Oracle Database Server.
      #
      - server: localhost:1522
    
        ## @param service_name - string - required
        ## The Oracle Database service name. To view the services available on your server,
        ## run the following query:
        ## `SELECT value FROM v$parameter WHERE name='service_names'`
        #
        service_name: "<SERVICE_NAME>"
    
        ## @param username - string - required
        ## The username for the user account.
        #
        username: <USER>
    
        ## @param password - string - required
        ## The password for the user account.
        #
        password: "<PASSWORD>"
    
        ## @param protocol - string - optional - default: TCP
        ## The protocol to connect to the Oracle Database Server. Valid protocols include TCP and TCPS.
        ##
        ## When connecting to Oracle Database via JDBC, `jdbc_truststore` and `jdbc_truststore_type` are required.
        ## More information can be found from Oracle Database's whitepaper:
        ##
        ## https://www.oracle.com/technetwork/topics/wp-oracle-jdbc-thin-ssl-130128.pdf
        #
        protocol: TCPS
    
  2. Update the sqlnet.ora, listener.ora, and tnsnames.ora to allow TCPS connections on your Oracle Database.

TCPS through Oracle without JDBC

If you are not using JDBC, verify that the Datadog Agent is able to connect to your database. Use the sqlplus command line tool with the information inputted in your configuration options:

sqlplus <USER>/<PASSWORD>@(DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCPS)(HOST=<HOST>)(PORT=<PORT>))(SERVICE_NAME=<SERVICE_NAME>)))

When using the Oracle Instant Client connection, move three files to the network/admin directory of the client libraries used by your application:

  • tnsnames.ora: Maps net service names used for application connection strings to your database services.
  • sqlnet.ora: Configures Oracle Network settings.
  • cwallet.sso: Enables SSL or TLS connections. Keep this file secure.
TCPS through JDBC

If you are connecting to Oracle Database using JDBC, you also need to specify jdbc_truststore_path, jdbc_truststore_type, and jdbc_truststore_password (optional) if there is a password on the truststore.

Note: SSO truststores don’t require passwords.

    # In the `instances:` section
    ...

    ## @param jdbc_truststore_path - string - optional
    ## The JDBC truststore file path.
    #
    jdbc_truststore_path: /path/to/truststore

    ## @param jdbc_truststore_type - string - optional
    ## The JDBC truststore file type. Supported truststore types include JKS, SSO, and PKCS12.
    #
    jdbc_truststore_type: SSO

    ## @param jdbc_truststore_password - string - optional
    ## The password for the truststore when connecting via JDBC.
    #
    # jdbc_truststore_password: <JDBC_TRUSTSTORE_PASSWORD>

For more information about connecting to the Oracle Database through TCPS on JDBC, see the official Oracle whitepaper.

For containerized environments, see the Autodiscovery Integration Templates for guidance on applying the parameters below.

ParameterValue
<INTEGRATION_NAME>oracle
<INIT_CONFIG>blank or {}
<INSTANCE_CONFIG>{"server": "%%host%%:1521", "service_name":"<SERVICE_NAME>", "username":"datadog", "password":"<PASSWORD>"}

Validation

Run the Agent’s status subcommand and look for oracle under the Checks section.

Custom query

Providing custom queries is also supported. Each query must have two parameters:

ParameterDescription
queryThis is the SQL to execute. It can be a simple statement or a multi-line script. All rows of the result are evaluated.
columnsThis is a list representing each column, ordered sequentially from left to right. There are two required pieces of data:
a. type - This is the submission method (gauge, count, etc.).
b. name - This is the suffix used to form the full metric name. If type is tag, this column is instead considered as a tag which is applied to every metric collected by this particular query.

Optionally use the tags parameter to apply a list of tags to each metric collected.

The following:

self.gauge('oracle.custom_query.metric1', value, tags=['tester:oracle', 'tag1:value'])
self.count('oracle.custom_query.metric2', value, tags=['tester:oracle', 'tag1:value'])

is what the following example configuration would become:

- query: | # Use the pipe if you require a multi-line script.
    SELECT columns
    FROM tester.test_table
    WHERE conditions
  columns:
    # Put this for any column you wish to skip:
    - {}
    - name: metric1
      type: gauge
    - name: tag1
      type: tag
    - name: metric2
      type: count
  tags:
    - tester:oracle

See the sample oracle.d/conf.yaml for all available configuration options.

Example

Create a query configuration to help identify database locks:

  1. To include a custom query, modify conf.d\oracle.d\conf.yaml. Uncomment the custom_queries block, add the required queries and columns, and restart the Agent.
  init_config:
  instances:
      - server: localhost:1521
        service_name: orcl11g.us.oracle.com
        username: datadog
        password: xxxxxxx
        jdbc_driver_path: /u01/app/oracle/product/11.2/dbhome_1/jdbc/lib/ojdbc6.jar
        tags:
          - db:oracle
        custom_queries:
          - query: |
              select blocking_session, username, osuser, sid, serial# as serial, wait_class, seconds_in_wait
              from v_$session
              where blocking_session is not NULL order by blocking_session              
            columns:
              - name: blocking_session
                type: gauge
              - name: username
                type: tag
              - name: osuser
                type: tag
              - name: sid
                type: tag
              - name: serial
                type: tag
              - name: wait_class
                type: tag
              - name: seconds_in_wait
                type: tag
  1. To access v_$session, give permission to DATADOG and test the permissions.
SQL> grant select on sys.v_$session to datadog;

##connecting with the DD user to validate the access:


SQL> show user
USER is "DATADOG"


##creating a synonym to make the view visible
SQL> create synonym datadog.v_$session for sys.v_$session;


Synonym created.


SQL> select blocking_session,username,osuser, sid, serial#, wait_class, seconds_in_wait from v_$session
where blocking_session is not NULL order by blocking_session;
  1. Once configured, you can create a monitor based on oracle.custom_query.locks metrics.

Troubleshooting

Common problems

Oracle Native Client

  • If you encounter a DPY-6000: cannot connect to database:
    Failed to connect to Oracle DB, error: DPY-6000: cannot connect to database. Listener refused connection. (Similar to ORA-12660)
    
  • Ensure Native Network Encryption or Checksumming are not enabled. If they are enabled, you must use the Instant Client method by setting use_instant_client: true.

For more information about setting up the Oracle Instant Client, see the Oracle integration documentation.

Oracle Instant Client

  • Verify that both the Oracle Instant Client and SDK files are located in the same directory. The structure of the directory should look similar:
    |___ BASIC_LITE_LICENSE
    |___ BASIC_LITE_README
    |___ adrci
    |___ genezi
    |___ libclntsh.so -> libclntsh.so.19.1
    |___ libclntsh.so.10.1 -> libclntsh.so.19.1
    |___ libclntsh.so.11.1 -> libclntsh.so.19.1
    |___ libclntsh.so.12.1 -> libclntsh.so.19.1
    |___ libclntsh.so.18.1 -> libclntsh.so.19.1
    |___ libclntsh.so.19.1
    |___ libclntshcore.so.19.1
    |___ libipc1.so
    |___ libmql1.so
    |___ libnnz19.so
    |___ libocci.so -> libocci.so.19.1
    |___ libocci.so.10.1 -> libocci.so.19.1
    |___ libocci.so.11.1 -> libocci.so.19.1
    |___ libocci.so.12.1 -> libocci.so.19.1
    |___ libocci.so.18.1 -> libocci.so.19.1
    |___ libocci.so.19.1
    |___ libociicus.so
    |___ libocijdbc19.so
    |___ liboramysql19.so
    |___ listener.ora
    |___ network
    |   `___ admin
    |       |___ README
    |       |___ cwallet.sso
    |       |___ sqlnet.ora
    |       `___ tnsnames.ora
    |___ ojdbc8.jar
    |___ ucp.jar
    |___ uidrvci
    `___ xstreams.jar
    

JDBC driver (Linux only)

  • If you encounter a JVMNotFoundException:

    JVMNotFoundException("No JVM shared library file ({jpype._jvmfinder.JVMNotFoundException: No JVM shared library file (libjvm.so) found. Try setting up the JAVA_HOME environment variable properly.})"
    
    • Ensure that the JAVA_HOME environment variable is set and pointing to the correct directory.
    • Add the environment variable to /etc/environment:
      JAVA_HOME=/path/to/java
      
    • Then restart the Agent.
  • If you encounter this error Unsupported major.minor version 52.0 it means you’re running a Java version that is too old. You need to either update your system Java or additionally install a newer version and point your JAVA_HOME variable to the new install as explained above.

  • Verify your environment variables are set correctly by running the following command from the Agent. Ensure the displayed output matches the correct value.

      sudo -u dd-agent -- /opt/datadog-agent/embedded/bin/python -c "import os; print(\"JAVA_HOME:{}\".format(os.environ.get(\"JAVA_HOME\")))"
    

Need help? Contact Datadog support.