by Porus Homi Havewala
We introduced the concept of Consolidation Planning in the previous parts of this article series. Part V was here. This is Part VI.
Consolidating to Virtual Machines
In the same way, we can create a new project for a P2V (Physical to Virtual) consolidation. This allows you to consolidate your existing Physical Machines to one or more Virtual Machines.
The consolidation project only applies to Oracle Virtual Machines. Consolidating to non-Oracle Virtual Machines, such as VMware, is not supported unless if you treat the VMware machines as Physical Machines and use the earlier P2P consolidation method..
When creating the project, simply select the Consolidation Type to be "From physical servers to Oracle Virtual Servers (P2V)". The other screens are mostly the same. You can select source candidates, then add existing virtual servers (current Enterprise Manager targets) as destination candidates. If you do not specify the destination candidates, only phantom (new) virtual servers will be used.
In the Pre-configured Scenarios step of the process, select New (Phantom) servers. This is seen in Figure 1-34. Note that there is no engineered systems option at this stage, however engineered systems can be chosen later on when creating the scenario.
Figure 1-34. Phantom Servers in the case of P2V Projects.
If you do not intend to use engineered systems as the destination, you can manually specify the CPU capacity, Memory, and Disk Storage of your phantom virtualized servers, along with optional entries for reserved cpu and reserved memory.
Finish creating and submitting the P2V Project.
Once the project is ready, you can create a scenario. In the case of a P2V scenario, a list of Exalogic configurations is provided, and you can select from that. We have selected the Exalogic Elastic Cloud X5-2 (Eight Rack) as shown in Figure 1-35.
Figure 1-35. Selecting an Exalogic X5-2 (Eight Rack) for the P2V Scenario.
The rest of the scenario calculations and mapping work in the same way, and the physical source servers are mapped to the destination phantom Eight Rack.
In this manner, the Consolidation Planner allows you to play a number of what-if scenarios, for P2P as well as P2V consolidations on the basis of sound mathematical calculations. You can specify which metrics will be analysed, and this results in the calculation of the resource requirements for every source server. Each resource is aggregated to a 24-hour pattern based on a different formula, depending on whether Conservative, Medium or Aggressive has been selected as the algorithm.
Constraints can also be specified as to which server workloads can co-exist together, and which workloads should be placed on different target servers for business or technical reasons.
Updating the Benchmark Rates
Now for a look at the Benchmark rates. The SPECint®_base_rate2006 benchmark is used by Consolidation planner for database or application hosts, or hosts with a mixed workload. The SPECjbb®2005 benchmark is used for middleware platforms.
The benchmarks are used as a representation for the processing power of CPUs - including Intel Itanium, Intel Xeon, SPARC T3, SPARC64, AMD Opteron, as well as IBM POWER.
The SPEC rates in Enterprise Manager can also be updated by users in the following manner. Go to the Host Consolidation Planner home page via Enterprise | Consolidation | Host Consolidation Planner, and select Actions | View Data Collection. On this page, see the section titled “Load Benchmark CSV File”.
The following information is seen on the page: “The built-in benchmark rates can be updated with a downloaded Comma Separated Values (CSV) file. To download the latest data, go to http://www.spec.org/cgi-bin/osgresults and choose either SPECint2006 Rates or SPEC JBB2005 option for "Available Configurations" and "Advanced" option for "Search Form Request", then click the "Go!" button.
“In the Configurable Request section, keep all default settings and make sure the following columns are set to "Display": For SPECint2006: Hardware Vendor, System, # Cores, # Chips, # Cores Per Chip, # Threads Per Core, Processor, Processor MHz, Processor Characteristics, 1st Level Cache, 2nd Level Cache, 3rd Level Cache, Memory, Base Copies, Result, Baseline, Published; for SPEC JBB2005: Company, System, BOPS, BOPS per JVM, JVM, JVM Instances, # cores, # chips, # cores per chip, Processor, CPU Speed, 1st Cache, 2nd Cache, Memory, Published. Choose "Comma Separated Values" option for "Output Format", then click "Fetch Results" button to show retrieved results in CSV format and save the content via the "Download" link in a CSV file, which can be loaded to update the built-in rates”.
As per the above instructions, the latest spec rates can be downloaded from SPEC.org in the form of CSV files, and then uploaded to the repository.
We will continue this article series in Part VII.
(This article is an excerpt from Chapter I of the new book by the author titled “Oracle Database Cloud Cookbook with Oracle Enterprise Manager 13c Cloud Control” published by Oracle Press in August 2016. For more information on the book, see here.)