Capacity planning is a key element in the management of virtualisation. It allows you to make the right investment decisions and anticipate project risks. We have designed capacity planning to be very intuitive and easy to use as soon as you feel the need: anticipate saturation dates, deploy new projects, simulate server losses, etc.
NOTE: The values displayed (USED, FREE, TOTAL) correspond to the state of each resource on the date displayed at the top left (state of the infrastructure on the last day of the filter).
Business Continuity Plan (BCP) and Disaster Recovery Plan (DR)
When running Capacity Planning simulations, you can select a priority state (High Availability) that will automatically select the virtual machines to be retained in the simulation.
Business Continuity Plan (BCP): All virtual machines on deleted servers are retained.
Disaster Recovery (DR): only virtual machines with the selected HA level are retained on the deleted servers.
Filter: only virtual machines selected in the filter are retained.
One of the advantages of virtualisation is resource over-provisioning: the virtual increase in physical resources (memory capacity, number of logical cores and disk space). For example: an infrastructure with a physical memory capacity of 100GB can virtually have For example, an infrastructure with a physical memory capacity of 100GB can virtually have a memory capacity of 200GB by over-provisioning the memory by 200%. By default, DC Scope uses the following over-provisioning rates: 4 VCPUs per logical core, 166% in memory and 100% in storage. More information is available in the Settings section.
Purpose: To provide saturation dates based on configurable criteria.
|Resource Allocations||Resource Consumptions|
|The left part named "Resource Allocations" corresponds to the allocated resources.||The right hand side named "Resource Consumptions" corresponds to the consumed resources.|
|VRAM: corresponds to the total amount of memory allocated to the VMs.||RAM: corresponds to the sum of the memories consumed ("consumed" counter) by the VMs.|
|VCPU: corresponds to the total sum of VCPUs allocated to the VMs.||CPU: corresponds to the sum of Mhz consumed ("cpuMHz" counter).|
|VDISK: corresponds to the total sum of disk space provisioned to VMs.||DISK: corresponds to the difference between the sum of the datastore capacities and the sum of the free spaces of these datastores.|
|USED: is the sum of the allocations for a given resource.||USED is the sum of the consumptions for a given resource.|
The saturation dates for each resource are determined from a linear regression. A colour code is used according to the date:
|Green||no short-term risk, (saturation date > 5 years)|
|Orange||near saturation (saturation date \< 5 years, month and year are displayed).|
|Red - KO||the resource is saturated|
|Red - KO||in the "Resource Allocations" section: it is not necessarily necessary to add physical resources. It may be sufficient to increase the overcommitment rate, resize oversized VMs or even remove idle VMs (to reduce allocations).|
|Red - KO||in the "Resource Consumptions" section: remove unnecessary VMs impacting resources and/or add computer resources.|
|Red - No data||Information not available.|
Add/Remove Project and Host Failure Modeling BCP/DR - What if scenarios?
The "Virtual Machines" section allows you to simulate the addition and/or deletion of VMs, based on existing profiles. By clicking on the "Deploy / Delete" button a window displays the inventory of VMs currently in the infrastructure. For each VM, we display its name, the amount of memory, the number of CPUs, the allocated disk space as well as an average of its memory (in MB) and processor (in MHz) consumption. The choice of VMs to deploy or delete is made by choosing the "deploy" or "delete" option next to their name. The chosen VMs appear then under the button.
Concerning the VMs to deploy, a pencil icon allows you to edit the metrics of the VM. It is also possible to to modify the quantity of VMs to deploy that will have this profile. Once the VM profiles to be deployed/deleted have been created, the simulation is launched by pressing the Once the VM profiles to be deployed/deleted have been created, the simulation is launched by pressing the "run simulation" button.
The simulation takes place in two stages:
- A placement algorithm will determine if the placement of the VMs is still possible. To do so, it will use the existing infrastructure on the last day of the filter and will simulate the addition of new elements or the deletion of existing elements.
- If placement is possible, we define new prediction dates taking into account the new capacities/allocations/consumptions of the elements (physical servers, virtual machines, datastores). In order to compare the prediction dates, the new dates related to the simulation are displayed below. The new saturation dates have the same colour code. The results of the simulation are available in PDF.
Host Failure Modeling
Objective: Simulate a hardware failure / room failover, Identify/Calibrate the hardware resources to be added.
Principle: The "Physical Machines" part allows the simulation of the addition and/or removal of physical servers from existing profiles. It is possible to carry out this simulation by selecting one or more servers and choosing all the VMs or only some of them (depending on the HA PRIORITY level chosen). For example, in the case of the loss of an entire room, this allows you to test whether the other room is able to host all the VMs that were on the "failed" servers or at least part of the VMs.
NOTE: It is possible to combine several simulations For example: "I have a new project of 'X' VMs, the simulation indicates that the capacity of my physical servers is insufficient of my physical servers is insufficient... I then simulate the addition of a physical server, ..."