Google Compute Engine (GCE)

Google Compute Engine is the service of lease of computing environments in a public cloud (IaaS) based on Linux OS providing services based on a payment for hourly resource consumption (computing powers and storages). Supports of Windows in these computing environments it is not stated yet. Let's remind that Google already has an App Engine PaaS-platform and service of cloud storage Cloud Storage. The service Compute Engine (CE) is available to trial use only by invitations (Limited Preview).

DPCs of Google are based in the central region of the USA, in Europe and in Asia.

Strengths: Google though late entered this market, but at the disposal of the company the computing resources and technical experience which do not have equal. These advantages will allow it to advance the Compute Engine quicker, than it will be done by competitors with the IaaS. Google "mainly turns already available opportunities into a product, i.e. the company does not need to create them from scratch", Gartner specifies a situation in 201.

What should be considered: Google is a courageous innovator in the field of purely cloud applicaions, but the company has no offer for the hybrid environment. Also, the company still should be credible in a corporate segment that will be not simply as Google not always provides migration of legacy workloads. And, still Compute Engine has no little significant program of partnership, specifies Gartner.

Main features of Google Compute Engine:

• Computing environments. Are available in configuration 1,2,4 or 8 of cores on Linux platform from 3.75 GB of RAM on one core.
• Storages. It is possible to store data of virtual machines in the form of non-constant (Ephemeral) and also permanent (Persistent) of disks on the party of Google or in Cloud Storage Google service. For permanent disks it is possible to do snapshots for the purpose of backup and also to provide access to one disk from several VM. All data of disks are ciphered.
• Network interaction. An opportunity to use high-performance network equipment of datatsetr of Google and to independently configure network screens, also to use external IP-adres. Plus ready-made solutions from third-party vendors which act as partners of Google.
• Management. Virtual machines can be operated via the web console or CLI, is also API giving a set of opportunities to developers and third-party producers.
• Google Compute Engine will be integrated with the partner solutions RightScale, Puppet Labs, OpsCode, Numerate, Cliqr, MapR and others.

In the spring of 2013 Google announced availability of Google Compute Engine (GCE) — IaaS- solutions which allow all subscribers of a packet of Google Gold Support to create and start virtual machines and to take all advantages of a virtual network. According to ZDNet, the company made the decision to create a system environment of GCE Debian GNULinux based on / instead of own Linux distribution kit Google Compute Engine Linux based on package base Ubuntu.

Now Debian will be used by software developers to default. Debian 6.0 'Squeeze' and 7.0 'Wheezy' will be supported. The last was let out just a week ago. Tools for deployment of private computing clouds are added to it. For simplification of deployment of own cloud infrastructures the structure of a distribution kit includes the packets allowing to set and customize with the minimum costs the cloud systems and infrastructures of virtualization based on OpenStack and XCP (Xen Cloud Platform). The possibility of use of Debian in cloud environments of Amazon EC2, Windows Azure and Google Computer Engine is provided.

GCE will continue to support a possibility of installation of images of CentOS 6.2. On The Register request concerning what else will be able to work as Linux distribution kits further with GCE, the representative of Google, without going into details, reported that 'the company on the basis of a feedback with clients continues to study actively opportunities for use of boot images and other operating systems'.

GCE allows to start virtual machines running Linux with 1, 2, 4 or 8 virtual kernels and 3.75 GB of memory on a core on demand. Virtual machines can be operated using the command line, the web interface or created by the user by means of API Google of management systems. The service is supported from many leading players in the field of cloud management, including RightScale, Puppet Labs, Opscode, Numerate, Cliqr and MapR. It should be noted that use of GCE which is a part of Google Cloud Platform costs 400 dollars a year.