Behind the longevity, should your server be changed?

What should the server update frequency be? This issue has been confusing with data center managers.

Fortunately, the duration of the server has been long. If new technology is used, the duration will be multiplied, and it will help save the company's funds.

Behind the longevity of the server?

First of all, thanks to server virtualization and hardware advancements, the server is able to maximize its benefits from consolidation and grow its life. Virtualization increases operational lifetime, especially as servers evolve to accommodate more processor cores and storage. Five years ago, a typical server had an average service life of about three years. In this case, companies often speed up the system and then upgrade to a more advanced hardware platform.

Server hardware upgrades include the introduction of virtualization extensions, such as AMD-V and Intel-VT. The processor has developed six-core, eight-core, ten-core, or even twelve-core, which alleviates the system. The evolution of the near-human relationship. The technology has reached 64-bit computing and now allows the server to support 128 GB, 192 GB, 512 GB or even 1 TB of RAM, which will extend the operating life of the system. The server also absorbs RAID-like rebound functions for local disk, multiple network adapters, and redundant power supply.

Clustering technology can prevent accidental downtime of old servers and relieve concerns for heavy-duty administrators. It can also avoid the old server failure dragging the whole world, they will have to be replaced once they are not working, after all, no one cares about the failure of the old server hardware, this node does not work on another node. The features that virtualization brings, such as live migration, allow workloads to quickly balance between servers and migrate applications from failed or over-loaded servers to new servers.

Service contracts are often expensive and become more expensive as server hardware ages. This became the main argument for supporting technology updates. Virtualization and clustering can alleviate most of this pressure in a production environment, but aging servers can bring tasks outside of data center products. For example, if the server has passed the time specified in the service contract, the system must be redeployed. Now, a common system can serve the business for six to eight years -- even until it is damaged.

How old is the old server?

But I have to give some warnings to long-lived server hardware. Perhaps most of the important issues are in parts availability. The server hardware will fail and the parts that need to be replaced are expensive. When a smart IT administrator redeploys an aging server, it considers the cost of the service and the availability of the part, makes a basic expenditure/benefit analysis, and then makes a more reasonable decision.

Economic considerations also affect the server update cycle. For example, a leased server must decide when to retire according to the schedule. In addition, companies may face higher rents when they rent. If you don't have a server, it makes sense to update the device within the mainstream rent.

Finally, advances in technology have made server updates necessary. For example, some outdated servers cannot carry a large number of virtual machines, which requires an update. In addition, powerful new processors, advanced processor expansion (especially related to virtualization) and energy-saving features will bring a new wave of server updates.