Here we start Designing server consolidation strategy .The term server consolidation typically means virtualization with technology such as Hyper-V,although that’s not always the case. Server consolidation can also mean consolidation of roles from multiple servers onto one server. When considering this latter form of consolidation, you must consider how consolidation of services will affect infrastructure reliability. For example, you don’t want to place too many services or roles onto one server in case that server goes down. Therefore, providing redundancy and separation of services to ensure uptime and reliability is key when designing server consolidation.
Server consolidation by virtualization involves several steps:
1. Determine the virtualization scope.
2. Create a list of workloads.
3. Select backup and fault-tolerance approaches.
4. Summarize and analyze workload requirements.
Determine the virtualization scope
The first step in designing a server consolidation strategy involves defining the scope of the project. For example, virtualizing an entire datacenter at once can be risky, as can the initial investment. Also, if the organization has multiple data centers, each might need to have virtualization hardware installed, which increases the initial cost of deployment. However, deployment to those remote locations can in fact provide good pilot projects for the overall virtualization deployment.
Create a list of workloads
Creating a list of workloads involves documenting what applications are currently deployed, their current location, what operating system those applications use, the resources requirements for the workload, and the administrator or person responsible for the application or workload.
The compatibility of the workload, specifically the ability for that application or workload to be virtualized, should also be undertaken at this time. If a given workload isn’t compatible or supported on virtualized infrastructure, it must be accounted for and excluded from the consolidation strategy. This task includes not only the technical compatibility but also the vendor’s support and licensing model compatibility when running in a virtualized environment.
DETERMINING RESOURCE REQUIREMENTS
Determining the resource requirements for the virtualized workloads is an important facet when creating a list of workloads. You should measure the current resources for physical machines and workloads, and then add resources to ensure that the load doesn’t meet its peak load immediately on being virtualized.
Table 1-1 lists three options for determining resource requirements.
TABLE 1-1 Options for determining resource requirements
Summarize and analyze workload requirements
So far you’ve collected a great deal of information about your organization’s infrastructure. This phase involves analyzing all that information to determine how best to virtualize the workloads involved.
In this phase, you look at how workloads can be grouped, keeping in mind that regulatory requirements might prevent workloads from being virtualized in the same datacenter or other requirements might deploy workloads at different locations.
Also during this phase, you gather actual resource requirements such as CPU, memory,disk (both capacity and performance), and network. The goal is to summarize those requirements so that you can plan what types and capacities are necessary to virtualize the workloads.
