Collectors

You can create collectors that run on the AppDynamics for Databases server to monitor any of the following systems:

• Databases: DB2, GreenPlum, Microsoft SQL Server, Microsoft SQL Azure Database, MongoDB, MySQL, Oracle, Oracle RAC, PostgreSQL, Sybase ASE, and Sybase IQ
• Servers: Linux Server, Solaris Server, and Windows Server
• Storage: NetApp and NetApp E-Series

Note: Monitoring Microsoft SQL Azure databases is accomplished through the Microsoft SQL Server Collector.

Add a Collector

1. Click Main Menu->Setup->Add Collector.
   
2.  Complete each field of the Add New Collector window and then click Add Collector.
    
3. After you add a collector, the new collector appears on the Collector Administration window. 

You will also now see your collectors on the Main Menu.

To change any of the collector attributes, click its name in the Name column of the Collector Administration window. The Edit Collector Configuration window appears where you can change the collector attributes except the Type and Name attributes.

The license is based on the number of concurrent databases monitored. If you configure more collectors than you are licensed to use, all your collectors will stop and you will need to delete collectors until you no longer exceed the maximum number of licensed instances.  A licenses message similar to the following displays:

Use the following to help you complete the fields of the Add New Collector/Edit Collector Configuration window. The fields that display depend on the type of collector you selected.

Complete the Metric Collector Section

Type: Select the collector type. There are three categories of collectors and collector types within these categories:

• Database Collectors: DB2, mongoDB, Microsoft SQL Server, MySQL, Oracle, PostgreSQL, and Sybase ASE
• Server Collectors: AIX Server, Linux Server, Solaris Server, or Windows Server
• Storage Collectors: NetApp, NetApp E-Series

Once you select the Type, the fields displayed may change because the information required differs between collector types. The type appears in the  Type column of the Collector Administration window.

Name: This name uniquely identifies the monitored instance as it appears in the Name column of the Collector Administration window.

• For versions prior to 2.9: Enter the name exactly as it is appears on the license.  
• For versions 2.9 and higher: Enter a name of your choice.
• For Microsoft SQL Azure, specify "SQLAzure" in the Name field.

Hostname or IP Address: Specify the hostname or IP address of the machine serving the database. The hostname or IP address appears in the Hostname column of the Collector Administration window.

Database: Specify the database name, named instance or SID of the database to monitor.

SID or SERVICE_NAME: (Oracle) Specify the services name or SID of the database to monitor. Click either SID or SERVICE_NAME to indicate which database identifier you are using.

Failover Partner: (Microsoft SQL Server) If you use Database Mirroring, enter the hostname of the IP address of the Failover Partner here.

EnterpriseDB?: (Postgre) Click if your Postrgre database installation is an EnterpriseDB distribution. 

Connect as a sysdba: (Oracle)  Click if you want to connect as user sys and enable password files. Running the collector using a sysdba account allows the collector to monitor an Oracle instance in standby mode, an instance that is used for failover which is being replicated from the main active instance.

Database Listener Port: Specify the TCP/IP address of the port the database uses for communicating with the AppDynamics for Database (App4DB) controller. The database list port appears in the Listener Port column of the Collector Administration window.

Username: Specify the name of the user AppDynamics for Databases uses to connect to the database. This user must have the Required Database Permissions. The user account used for monitoring a SQL Server database can be a Windows authenticated account (if the AppDynamics for Databases is running on Windows) or SQL Server authenticated (if AppDynamics for Databases is running on Windows or Linux). 

Password: Specify the password of the user AppDynamics for Databases uses to connect to the database.

Aggregation Interval: Specify the time interval at which AppDynamics for Databases should write out the performance data to the repository. Shorter aggregation intervals provide more granularity and closer to real-time access to data; however, more disk space is required to store historical data and there will be a small increase in overhead as the target system is queried more frequently. For more information, see Configure the Aggregation Interval.

Logging Enabled: Click to enable verbose mode logging, which logs all communications between the controller and the collector. Enable only during troubleshooting because logging can consume a lot of disk space. If you have enabled logging, you can click the logging icon in the Log column of the Collector Administration window to view the log file. The log files are located in the <App4DB install directory>\agent directory and have the format CollectorName_out.log and CollectorNam_err.log. 

Complete the Host Collector Section

Monitor OS?: Click if you want CPU consumption metrics collected from the monitored host. For more information, see Infrastructure Monitoring (Monitor OS?).

OS Type: Specify the operating system of the monitored host: Linux, Solaris, or Windows.

Hostname: Specify the hostname or IP address of the monitored host.

SSH Port (Linux/Unix Only): Specify the Secure Shell (SSH) port number AppDynamics for Databases should use for encrypted communications with the monitored host. The default port number of 22 will be used if you do not specify a different port number here.  AppDynamics for Databases also supports certificate-based authentication via Privacy Enhanced Mail (PEM). To implement certificate-based authentication, copy the PEM file to the <AppD4DB install directory>\agent directory.

Username: Specify the name of the user AppDynamics for Databases uses to log on to the monitored host.  To collect OS metrics from a Windows host, the configured user (or Collector Service user if using Windows Authentication) must be able to establish a WMI connection to the target host and collect Windows Performance Counters.

Password: Specify the password of the user AppDynamics for Databases uses to log on to the monitored host. The number of echo characters shown in the password text field should not be interpreted to imply the number of characters stored for the (encrypted) user password.

Database Specific Collector Configuration

Configure a Collector for MongoDB

You only need to configure a MongoDB collector for any one node of the replicaset (either a primary, or any secondary) as AppDynamics for Database will automatically discover all other nodes. To configure the AppDynamics for Databases collector you will need the IP address and the TCP port number of the mongod daemon.

For a Sharded environment, you only need a collector that connects to the mongo daemon. AppDynamics for Databases will discover all of the replicasets in the cluster based on that one connection, and shouldn’t have connections configured for each replicaset.

Configure a Collector for a Microsoft SQL Server Cluster

In a SQL Server cluster environment with two virtual nodes, primary and master, configure the AppDynamics for Databases collector with the cluster IP address and the TCP port number of the SQL Server instance. The DBA can provide you with the instance IP address and port address, but if not you find this information in the Microsoft SQL Server Configuration Manager GUI.

Configure a Collector for Microsoft SQL Azure Database

Configure the collector for the Microsoft SQL Azure Database as you would the Microsoft SQL Server, with the following exceptions:

• Specify "SQLAzure" in the Name field.
• Host monitoring is not supported.

Configure a Collector for Oracle RAC

To monitor an Oracle Real Application Cluster (RAC), you must configure a collector for each instance of the RAC cluster. On startup, the Oracle collector detects whether it is monitoring a RAC or a single-instance database. If the collector detects a RAC, it finds the other instances in the RAC and combines the metadata for both instances.

The collector setup is as normal, with the exception of the Oracle SID which is often confused in an Oracle RAC set up. Whereas normal RAC clients connect to the cluster via the service name, AppDynamics for Databases needs to connect to each node individually via the local instance name. To find out the local instance name you need to connect via sqlpls (or equivalent) and run the following command:

SQL> show parameter instance_name;

Configure a Collector for a Database on Amazone RDS

See How to Set Up and Monitor Amazon RDS DatabasesAppDynamics for Databases.