All about Google App Engine

App Engine Google Services of Hosting Sites and web-applications on servers Google with free name SiteName.appspot.com, either with his own name, involved using the services of Google.

App Engine is presented in April 2008, are available as free accounts ... up to 1 GB of disk space, 10 GB of incoming traffic on the day, 10 GB outbound traffic per day, 200 million gigabyte the CPU in the day and 2,000 operations, send e-mail day, and the possibility of acquiring additional resources.

Applications, takes place on the basis of App Engine, you must be written in Python, Java, Go or PHP. The latter two still have the status of experimental instruments, although Google and vows to expand the list of supported programming languages and environments, and the service do not depend on any programming language. Runtime Python includes the full realization of possibilities of Python, most functions of standard library, a limited version of Django, and so on.

A set of API for storage services, datastore API (BigTable), Accounts Google, data load on the URL, E-mail, and so on.

The opportunity to use the scheduler cron for applications implemented on Python, and in Java. Allowed planning is not more than 20 jobs.

Using Google Accounts service allows you to quickly start using the application, there is no need to carry out a separate registration credentials on each site. It also allows the developer does not take care of the implementation of another user registration system specifically for your application.

Google's platform competes with similar services from Amazon, which provide opportunities to place files and web applications using their infrastructure.

Unlike traditional hosting model application

Unlike many conventional hosting applications on virtual machines, such as Amazon EC2, App Engine platform is tightly integrated with applications and imposes certain restrictions on developers. Environments parallel processing allow the use of much of the software created for Unix-like systems, while App Engine requires the developer to use one of the programming languages Python, Java, Go (since May 2011) or PHP (experimental), and used to store information storage (Datastore) - a subset of a proprietary database BigTable (en), which is the know-how of the company Google.

Although many of the developed language Python application may run without changes to App Engine (excluding those that require access to low-level functions of the system or access to the network), the use of storage require considerable forces developers to adapt applications. In contrast to popular databases MySQL and PostgreSQL, storage is not a relational database, does not require schema and data normalization. The platform architecture is scalable without changing the code. Implementing support for relational logic would lead to a significant slowdown. Non-relational nature of storage requires developers to use non-traditional storage architecture and data processing.

At the same time, these restrictions can lead to the fact that many developers will have to make additional efforts to adapt to the new environment, and they would prefer the choice of alternative platforms.

Limitations

The runtime platform Google App Engine enables developers to control only a part of the operating system, which can be a problem for application development.

No write access to the file system server. The only way to save your data - internal storage, non-relational, highly scalable database. Storage is different from relational databases MySQL and PostgreSQL, which are used on most sites.

Session management systems that use the file system will not operate. Google provides an alternative option - Memcache API. There is a possibility to organize a user authentication accounts through Google.
Google is widely used in the applications technology MapReduce, and its support will likely be added to the capabilities of the platform.

App Engine is SQL support SQL Cloud with Google.