Google and Parallels Agree: Container Technology Is the Best

Containers – The Best Choice for High Density, High Performance Cloud Virtualization

As the Google example below demonstrates, Containers have significant advantages in applications such as Cloud VPS hosting that rely on maximizing hardware efficiency while at the same time delivering maximum performance.  Containers are used by the majority of Service Providers delivering Cloud VPS hosting services today to deliver: 

• Unmatched density
• Near-native performance
• Dynamic resource control
• Complete  isolation
• Built-in application provisioning
• Substantial disk space savings
• Substantial memory savings

To learn more visit http://www.parallels.com/products/pvc/ 

Google’s Data Center Provides Latest Example of Advantage of Containers

Virtualization of physical servers has rapidly become ubiquitous across most industries.  There are two types of virtualization commonly available on the market today. Hardware virtualization, or hypervisor, virtualizes at the hardware level creating a duplicate of all system resources and resulting in overhead on the server and reducing the ratio of virtual servers per physical server as well as overall system performance.

Containers virtualize at the host operating system and provide a common virtualization layer that allocates system resources across all virtual servers, called Containers. The result is a leaner, more efficient virtualization layer that translates into higher ratios of virtual servers to physical servers, near-native server performance and lightning fast management operations.

Google Data Center Demonstrates Containers Advantages

At the recent Linux Plumbers Conference, Google presented their data center architecture and discussed why they rely exclusively on Containers as their virtualization technology of choice and how Containers allow them to pack far more of their jobs onto physical systems than they could using hypervisors.

Google uses Containers to run a two tier data center and maximize the efficiency of their server hardware.  Tier 1 jobs include customer visible and response critical tasks like search and ad-word display while Tier 2 jobs are tasks that require compute power but don't have a fixed deadline, like analytics and map-reduce operations. 

Google relies on the fast response of Containers to enable the system to respond almost instantaneously to increased capacity requests from tier 1, enabling maximum hardware usage without creating visible latency in search result display time. 

For more information, the presentation can be viewed online here: http://www.linuxplumbersconf.org/2011/ocw//system/presentations/705/original/Control_Loops_In_Userspace.pdf