Product and service reviews are conducted independently by our editorial team, but we sometimes make money when you click on links. Learn more.

Getting Started with the Cloud: Google Compute Engine

Getting Started with the Cloud: Google Compute Engine

Getting Started with the Cloud: Google Compute EngineGetting Started with the Cloud: Google Compute EngineWhen it comes to running large data centers distributed around the globe, Google is at the forefront. Amazon AWS may be almost synonymous with public cloud computing but Google has pushed Web-scale technologies like no other - from commodity data center architectures to scalable SQL and NoSQL data stores, Google has created and shared cloud technologies

Google could be the Amazon of cloud computing but it is not, at least not yet.

Google Compute Engine and related data storage services are Google’s foray into public IaaS. The current version of Google’s general purpose IaaS offering is in limited preview and well suited for batch, data intensive analysis but it does not have all the features you may have come to expect from a one-stop shopping public IaaS provider. If you are evaluating public IaaS vendors for big data, data warehousing, high performance computing or other analytics- focused applications, then keep Google on your short list. 

For readers looking for an IaaS provider that can scale your Web site automatically and load balance across multiple transaction processing servers, then you should consider others. Google Compute Engine is still in limited release so it does not have all the features you might be looking for at this time. With this article, Windows Azure becomes the latest platform we cover in our ongoing IaaS Buyer's Guide.

Getting Started with Google Compute Engine

Google asks customers to provide a description of how they will use Google Compute Engine when they sign up.  If a workload fits Google’s criteria then the customer is granted access to the Google cloud. The Google IaaS offers a few Linux distributions, including Ubuntu and CentOS, but no Microsoft Windows Server operating systems. Virtual machines are grouped into three categories: standard, high memory and high CPU.

Standard VMs range from one virtual core with 3.75 GB of RAM and 420 GB of local storage to eight virtual cores with 30 GB of RAM and two 1770GB disks. Prices start at $0.138/hour for a one virtual core machine hosted in the United States to $1.04/hour for an eight-core virtual machine. Prices are slightly higher for virtual machines running in European data centers. If you do not require local storage (other than the boot disk), diskless options are available from $0.12/hr to $0.96/hr for U.S. hosted virtual machines.

High memory machines are available in two-, four- and eight-core configurations with 13GB, 26GB, and 52GB of RAM, respectively. Prices ranges from $0.254/hour for a two core diskless VM to $1.72/hour for an eight-core VM with two 1770GB disks for local storage.

High CPU configurations are available with two, four and eight virtual cores and 1.8GB, 3.6GB and 7.2GB of RAM respectively. The two-core diskless High CPU configurations costs $0.136/hr; the eight-core configuration with two 1770GB disks for  local storage costs $0.68.

Data transfers into the Google Compute Engine cloud are free as are transfers to other cloud services in the same region. Data transfers out of the Google cloud to a destination in the Americas, Europe, Middle East or Africa are $0.12/GB for the first terabyte dropping to $0.08 for total monthly transfer over 10 terabytes. Data transfers out of the Google cloud to Asia Pacific locations are billed at $0.21/GB up to 1 terabyte and dropping to $0.15/GB for total monthly egress over 10 terabytes.

Data stored on virtual machine disks are ephemeral and lost when a virtual machine is stopped. Google Compute Engine has two options for long-term storage: persistent disks and Google Cloud Storage. Persistent disks offer performance comparable to local disks. Data is replicated to multiple devices within a data center to improve durability and availability. Persistent disk storage is available for $0.10 GB/month for general-purpose storage but snapshot storage is higher at $0.125 GB/month. In addition, I/O operations are charged at a rate of $0.10 per million. Google Cloud Storage is an object data store that offers standard and reduced availability options. Prices for standard object storage start at $0.085 for the first terabyte and drop to $0.054 when storing over 500 TB. Reduced availability storage prices range from $0.063 to $0.042.  The reduced storage service is currently considered experimental so the Google service-level agreement (SLA) does not apply to it.

Customers can manage their cloud resources using either a management portal, which is part of the Google API interface, or through a command line tool.  The management portal also provides access to other Google cloud services such as Google Cloud Storage and Google Cloud SQL.

Google’s service-level agreement is short and to the point. Services are operational and available to customers 99.95% of the time or the customer will receive a financial credit on a sliding scale. When services are available 99.00% to 99.95% of the time, the customer can receive a 10% credit. Customers receive a 25% credit when the service is available only 95% to 99% of the time and a 50% credit if the service is available less than 95% of the time.

Dan SulivanDan SulivanDan Sullivan is an author, systems architect, and consultant with over 20 years of IT experience with engagements in systems architecture, enterprise security, advanced analytics and business intelligence. He has worked in a broad range of industries, including financial services, manufacturing, pharmaceuticals, software development, government, retail, gas and oil production, power generation, life sciences, and education.  Dan has written 16 books and numerous articles and white papers about topics ranging from data warehousing, Cloud Computing and advanced analytics to security management, collaboration, and text mining.

See here for all of Dan's Tom's IT Pro articles.