VMware vCenter Data Acquisition Jobs

Although staging jobs for SAS IT Resource Management adapters are typically run once a day, the VMware vCenter adapter requires the ability to run staging jobs more frequently than once a day. The reason for this frequency is that staged VMware data must include information about the mapping of a host (physical machine) to the guest (virtual machine). As virtual machines are moved from host to host, the VMware database is updated with the current location of the virtual machine. The VMware database stores only the most recent host-to-guest mapping information. It does not store any previous mappings. Therefore, if the host-to-guest mappings change more than once a day, those mappings must also be extracted from the VMware database more than once a day. Those mappings must also be stored, along with some of the corresponding data, in a table for use as source tables in the nightly run of the VMware staging transformation. The VMware vCenter Data Acquisition transformation provides this functionality.

Like all other transformations, the VMware vCenter Data Acquisition transformation is contained in a job. When the VMware Data Acquisition job runs, it extracts a subset of VMware data from a database. The target table that is created by executing the VMware vCenter Data Acquisition transformation is input to the VMware vCenter staging transformation. Unlike the staged tables created by other staging transformations, the target table is not purged when the VMware Data Acquisition job is executed. The VMware vCenter Data Acquisition staging transformation prompts for the following information:

The target table that is created is a staged table. It should be used as input to the VMware vCenter staging transformation. When this staged table is input to the VMware vCenter staging transformation, it is merged with other VMware tables by the VMware vCenter staging transformation during the nightly staging run for that adapter.

The VMware Data Acquisition staged table has its own library. The Adapter Setup wizard creates the two staging libraries—one for the VMware vCenter Data Acquisition table, and the other library for the other VMware vCenter staged tables. (If created by the Adapter Setup wizard, these tables and libraries reside in the Staging folder of the IT data mart. If created manually, outside of the Adapter Setup wizard, these tables and libraries reside in the location of your choice.)

The VMware DataAcquisition staged table that is output from the VMware Data Acquisition staging job contains useful information about the topology of the VMware environment at the time that the VMware Data Acquisition staging job is run. The topology of the VMware environment includes the names and IDs of each guest, and the host or cluster that each guest belongs to. However, the VMware DataAcquisition staged table does not preserve the information about the movements of a guest from host to host or cluster to cluster between staging runs.

If you want to preserve this information over a period of time, you can instruct the VMware Data Acquisition staging job to create a separate data set to contain this information—the VMwareTopologyDimensionTable. To specify the creation of that table, set a macro variable called CreateTopologyDimensionTable to YES in the generated code (or deployed job code) for the VMware Data Acquisition staging job. To specify that the VMwareTopologyDimensionTable be created, add the following line to the generated code for the VMware Data Acquisition staging job:

%let CreateTopologyDimensionTable = YES; 

The table will be stored in the same staging library as the DataAcquisition staged table that is output from the VMware Data Acquisition staging job.

You can use this table as required at your site. The table is not used by the VMware staging code.

The VMware Data Acquisition job should be scheduled to run whenever it is necessary to capture the movements of virtual machines across physical machines in your VMware environment. If your environment changes rapidly, then the VMware Data Acquisition job might need to run as frequently as once an hour. However, if the virtual machines in your environment do not move between hosts as often, then the VMware Data Acquisition might need to run only once per day. This scheduling is dependent on, and should be arranged according to, the needs of your site.