Changes brought about by the rise in cloud computing have had a huge impact on servers and server design. Unlike enterprise IT departments, cloud companies like Amazon, Google and Facebook tend to custom design their infrastructures to handle various types of workloads in the most efficient and economical ways.
One such category of workloads includes light computing tasks with lots of parallelism and frequent use of large amounts of data. That is where Microservers come in…
Microservers are designed to process lightweight, scale out workloads for hyper-scale data centres. Typical workloads suited for Microservers include static web page serving; entry dedicated hosting, and basic content delivery, among others. Because of the Microservers’ high-density and energy-efficient design, its infrastructure (including the fan and power supply) can be shared by tens or even hundreds of physical server nodes.
Key Attributes of a Microserver
Microservers are generally based on small form-factor, system-on-a-chip boards, which pack the CPU, memory and system I/O onto a single integrated circuit. Power consumption of Microserver boards is much less than that of the processors in a typical high-end server.
(2) Space Savers
Because of their small size, and the fact Microservers require less cooling than their traditional counterparts, they can also be densely packed together to save physical space in the datacenter.
Since Microservers typically share infrastructure controlling networking, power and cooling, which is built into the server chassis, hence compared to alternatives, can cost less to run.
Microservers are typically suitable for static web page serving; entry dedicated hosting, and basic content delivery. They also lend themselves to data processing tasks where the workload can be parceled up and operated on in parallel, such as certain analytics jobs (think Hadoop), as well as handling data in non-relational databases.
(1) Not Suitable for compute intensive jobs
Because of their relatively lower compute performance or memory footprint, Microservers may not be suitable for handling mainstream enterprise IT or advanced scientific or technical computing workloads.
(2) Limits on parallelism
Due to current challenges and cost of adapting software to efficiently distribute workloads between Microservers; Microservers will require software to be rewritten to enable these servers to process the tasks in parallel.
(3) Networking overheads
Deploying a large number of less powerful servers increases the number of ports required and switching overhead.
Facebook is putting ARM and Tilera processors through tests in its datacenters, as it explores using non-x86 chips for some applications to reduce its electricity bill and boost performance. Dell and HP (Project Moonshot) are coming up with their Microserver line of products. CPU wars between Intel and ARM are heating up in the Microserver area.
Microservers are going to be specialized and custom-built to handle specific use-cases and will have their own place in the data-centers, both in enterprise IT and hyperscale data-centers. They will sit alongside rather than replace traditional higher power, less specialized servers.