Find the best server consolidation rate

How much is the server consolidation rate appropriate? Is there a risk to high server consolidation rates?

The important fact that a novice needs to know is that server consolidation is not a single static number--without the right rate of server consolidation, each enterprise must identify the sensible integration goals that are appropriate to their unique requirements. For example, 100% of the consolidation is certainly possible (all available server computing resources are used), but this is not necessary or appropriate.

If there are 100 physical servers in an organization. Assume that these servers are virtualized and that each server is carrying a workload. This is the 2:1 server consolidation rate, which reduces the server demand by half. 50 servers can complete 100 server workloads. It also cuts half the server's money and every month's electricity.

Workload migration software, such as Microsoft's dynamic migration, is a common tool for moving and consolidating workloads. As experience and confidence in virtualization grow, organizations typically implement the subsequent phases of consolidation as resources permit, systematically migrating more workloads to fewer servers. For example, the 2:1 integration mentioned above will be further consolidated later and may save another 10 servers.

In practice, all (100%) integration is undesirable. For example, some workloads may present a high variable or cyclical resource requirement. This requires that additional resources be temporarily allocated to the workload for the day or month of the day or for a higher period of user demand. In other cases, you should carefully reserve some computing resources for each server to supply failover to other servers. If each server is fully consolidated, there will not be sufficient resources for failover, which may cause some of the affected workloads to be unavailable unless the problem server is resolved.

Higher consolidation always poses greater risks, such as load availability and recovery time.

Suppose you have a VM on a virtualized server. If the server goes wrong, IT staff only needs to restore or restart the workload on another system. If multiple VMS use the same system, the problem is a bit tricky. When a VM is started, it uses bandwidth and computing resources-the bandwidth and computing resources that are left for the second VM to restart, and fewer resources to restart the third fourth VM. For a system with more than 10 workloads, a thorough recovery may take a considerable amount of time.

As a result, organizations must address and mitigate the reliance on fewer server platforms. In addition to increased network bandwidth and storage I/O performance, next-generation servers typically have superior reliability, such as redundant power supplies or memory hot backups. Other methods include server clusters or VM replication. Regardless of the method, the goal is to prevent workload outages for the first time, thereby alleviating the fix problem.

In general, management software that supports virtualization systems can help each workload on any server get the appropriate resource utilization. IT administrators can make informed decisions about VM migrations and find the optimal rate of server consolidation.