Evolution Hosting's technical architecture provides a robust environment for enterprise websites. The environment allows for powerful application management and monitoring. It also provides the ability to move and provision new resources as accounts change and grow.

Figure 4.1 represents a high-level view of the Evolution Hosting environment. A customer account is shown along with common services, including DNS, e-mail, and backup. The account is fronted by a firewall and all servers are controlled by myEvolution agents.

Accounts can be custom provisioned to exactly match customer requirements. Either shared or dedicated platforms can be requested by account holders to best fit budgetary and performance needs.

Evolution Hosting's technical architecture is designed for reliability and manageability. The environment provides monitoring, application management, resource provisioning, and automated corrective action capabilities to all accounts.

Figure 4.2 illustrates a typical Evolution Hosting account setup with the myEvolution hosting control center.

A major feature of the architecture shown in Figure 4.2 is the separation of the management area from the production servers. myEvolution has all of the information it needs to recreate any of the production tiers on new physical servers in the event of a catastrophic failure . The net result is increased reliability for Evolution Hosting customers.

Account holders use their myEvolution to deploy applications to a working area. From this location, the application needs to be pushed out to the account's live production servers. Content is pushed using a publish process.

The Publish process is more correctly a synchronization process due to the fact that content can move both directions between myEvolution and the production environment. When the same content exists in both places, the one with the most recent modified date is used as current. The current content item is then copied to the opposite server, overwriting old content in the process.

Figure 4.3 illustrates the publish process pushing files out to a live customer account.

The recommended method of moving content to Evolution accounts is first to upload content to the myEvolution hosting control center and then to Publish it to the production environment.

There is an optional Direct Upload process that account holders can use to push content directly to production environments. The Direct Upload eliminates the extra step of clicking the publish button in myEvolution.

The advantage to using Publish versus the Direct Upload method is that there is an instant backup of content in the myEvolution control center. If an account holder wishes to switch between shared and dedicated packages, or if an error occurs in the current physical server, the entire account can be immediately restored to a new server.

Content placed directly into the production environment via Direct Upload will be synchronized to myEvolution only after a Publish command is given or when one of the servers is restarted by the account holder from inside myEvolution.

Instructions on executing the publish command are available in the section called “Publish”.

In the event that a catastrophic failure occurs in an Evolution account's application or database, Evolution has put in place a comprehensive set of technologies and procedures that allow a quick and complete recovery. This system consists of hardware, software, and remote storage appliances that provide for the backup, analysis, and restoration of an account.

Evolution's disaster recovery system allows for either coarse or fine-grained restoration. Backups and Evolution Administration tools can be used to recreate an individual user's mailbox or an entire database server. The overall suite of disaster recovery procedures translate into increased uptime and reliability for account holders.

The details below are a minimum feature set that every Evolution account has. Dedicated hosting has additional options above and beyond the list below. This may include changes to the frequency of backups, application or server clustering, and other recovery measures that an account holder may wish to implement. Please contact <sales@evolutionhosting.com>. to discuss specific needs in further detail.

The Evolution hosting platform is built from the ground up with redundancy in mind. Some examples of this include:

Evolution Hosting has built its systems to continue functioning in the event of a catastrophic failure.

Evolution maintains hot-standby servers that are preconfigured with the myEvolution hosting environment and remote agent. Should an account's server become unusable, Evolution staff can use a myEvolution administration console to dynamically move the account from the corrupt server to a new hot-standby server. This feature is automated and takes only a matter of seconds.

RAID stands for "Redundant Array of Independant Disks". RAID allows a server to continue functioning if one or more hard disks fail. Each server in the Evolution environment has RAID and therefore has increased reliability at the disk-hardware level.

This section contains information on each type of backup that Evolution monitors and runs for account holders. One important metric with regard to backups is the frequency with which they run. Table 5.1 illustrates the schedule on which Evolution backup tasks occur.

* Configuration changes and content are synchronized between a myEvolution Server and the Customer Application Server.

Evolution has created a number of tutorials to assist account holders with the development, testing, and deployment of applications to myEvolution. The tutorials are one of the cornerstones in the Evolution Hosting system and it is highly recommended that each account holder download and take advantage of the tutorial that corresponds to his or her account.

The tutorials generally correspond to a particular application server. As vendors release new application server versions, Evolution staff update or create new tutorials to best fit the new offering.

Although each tutorial is different, they all are based on the same design and have similar layers of functionality. These layers are discussed in the next section.

To understand how the turorials work from a high-level, the tutorials can be divided into layers. Each layer has a different purpose:

The Domain Name System (DNS) is a physical address lookup utility that has evolved as a core piece of the Internet's operation. The DNS allows Internet addresses in mnemonic form such as www.companyx.com to be resolved into a corresponding numeric IP address such as 134.220.4.1 . It is this numeric address that allows computers to locate one another.

To ensure that addresses are not changed by anyone but the registered owner, the DNS requires that addresses be changed through a Registrar. Evolution takes care of a majority of DNS configuration, but the Registrar changes must be done by the customer.

This section of the Evolution manual explains when and how to change DNS settings so that Evolution accounts can be located on the Internet.

When any account is signed up via the Evolution Hosting automated signup process, two (2) domains are instantly set up within the Evolution DNS infrastructure. Evolution calls these domains the primary domain and the auxillary domain.

The primary domain will usually be the domain customers use to access the site. It will not be available, however, until the account holder completes the DNS Server Name Change process with their Registrar. Prior to the DNS change, the auxillary domain is used by the customer for immediate access after signup and for initial testing. In the following example a primary domain and auxillary domain are listed. Note that the auxillary domain is simply the primary domain with .evohst.org appended to the end of it.

Example Domain Set

Type: Primary Domain

Example URL: www.companyx.com

Derived From: Entered by customer during account signup

Use: Used by site visitors

Available: Only after account holder updates DNS settings with their registrar

Notes: Evolution accepts all valid domain names, including International domains

Type: Auxillary Domain

Example URL: www.companyx.com.evohst.org

Derived From: Created by myEvolution by appending .evohst.org to the primary domain.

Use: Used by Evolution staff and account holders

Available: Available immdeiately after account signup

Notes: The auxillary domain is not valid for e-mail address use - it is intended for web browser use only.

The primary domain is not available for use until the account holder changes the Internet's central DNS records. This change causes the Internet's global DNS system to point to the Evolution DNS subsystem. The end result is that when a URL is typed into a web-browser anywhere on the Internet, the browser will successfully find the Evolution hosted web site.

The Internet's DNS system has a time lag of up to 72 hours (three days) in propogating DNS updates to all DNS servers around the world. In cases where a previously existing site is active, it is advisable to keep both the previous site and the new Evolution site live during the 72 hour transition period. If the previous site is turned off before 72 hours has passed some customers may still be sent to the old site and receive an error.

Evolution has two (2) DNS servers that must be communicated to registrars for each new Evolution account. Please note that Evolution servers should not be used by your own network. Please be sure and use the DNS servers provided by your local internet access provider.

The two subsections that follow give step-by-step instructions for updating DNS records with a few of the Intenet's leading registrars.

A standard myEvolution account will have a single domain name that points to a single set of content. Some accounts, however, require multiple domain names and have multiple sets of content that correspond to the different names. Domain Parking is the term that Evolution uses to describe the process of working with more than one domain name in the context of a single account.

An example of domain parking is as follows: Company X may have a primary domain name of www.companyx.com, but may have another domain name of internal.companyx.com and require that they point to two different applications within their account. Using properties within myEvolution, account holders may set up their account to forward requests for www.companyx.com to one set of application content, while requests bound for internal.companyx.com are forwarded to a second set of application content.

Note: please also read the DNS section of this manual for more information related to domains, URL's, and IP addresses.

The following sections illustrate various domain parking scenarios, explaining both when they should be used, and how to configure them.

Account holders sign into myEvolution via the login page. A username and password are required. If an account holder misplaces their username or password, they may use the reminder feature on the same page to have login credentials e-mailed to the account's primary contact.

Account Holders access the various sections in myEvolution by clicking on the primary and secondary navigation links at the top of the page. These links utilize JavaScript and HTML Frames to optimize performance. The primary navigation links (top row) dictate what is displayed in the secondary links (bottom row). For example, once the database tab is clicked on, the bottom row of links is updated to display the SQL Tab, the Import Tab, etc.

The Server Control is located in the upper right corner of myEvolution and is used to control the life cycle of an account's application server.

To start, stop, or restart an account's application server the Server Control Buttons are used.

myEvolution has context-sensitive help in every section Help appears to the right of the navigation links and contains both a section description and a link to the user manual for more information. For additional help, the Support section of myEvolution contains links to the user manual, Evolution Support contact form, and priority electronic paging system.

The Deploy Section of the myEvolution Hosting Control Center is one of the most frequently accessed areas in the Evolution environment. From Deploy, account holders can upload new application content, validate EAR's or WAR's, and Publish applications.

For a quick-start to deploying applications to myEvolution, skip ahead.

A majority of the Deploy section is devoted to uploading files to directories in the Evolution environment. It is helpful to view Chapter 11 for reference on the underlying directory structure behind the various myEvolution Deploy sections. Account holders should also read the Publish section before working in the Deploy section to understand the underlying architecture of myEvolution and its method of delivering applications to production servers.

The remainder of this chapter explains the Deploy subsections in detail.

This section gives tips and abridged deployment instructions to deploy a J2EE WAR into the Evolution environment.

The JAR's section of myEvolution allows multiple Java Classes to be deployed in aggregated form.

To deploy JAR files to myEvolution:

The Classes contained in the uploaded JAR files are automatically added to the Classpath of the pertinent Classloader(s) for the account. Exact classpath settings for an application can be viewed in the section called “Startup Log”. Account holders should note that myEvolution includes all core classes required by various application servers and these files should not be uploaded into the JAR's Section or ClassCastExceptions may result.

The Data Section can be accessed by applications and is specially configured to not be accessible via site visitor's web browsers. An example use for this tab would be to store signup confirmation files generated by an application. The files could later be downloaded via the Data Section for review. To see how applications access the Data directory and related properties see Chapter 9

The FTP Section contains reference information and credentials explaining how to transfer files to and from an account via FTP. Figure 4.3 illustrates FTP upload options.

The Publish Section has one link, which activates the Publish routine for an account. Read more about Publish in the section called “The Publish Process”.

The Server section of the myEvolution Hosting Control Center contains logs and property information related to an account's servers and other resources such as DNS.

The logs section should be consulted frequently by account holders to check server startup results and application output.

The properties section contains server properties and other settings for various aspects of the Evolution Hosting environment. A description of all properties can be found in myEvolution Properties Glossary.

Each Evolution Hosting account has at least one database. Evolution databases are primarily used by an account's Java application, but may also be used by other applications (e.g., cgi-bin). This section provides information on the general setup and management of databases, as well as instructions on how to connect to a database from within an Evolution hosted Java application.

For answers to frequently asked questions, visit the Evolution Hosting Databases.

Each account's database tablespace is created automatically by the myEvolution Hosting Control Center upon account signup and is immediately ready for use. Account holders do not have to do anything extra related to the tablespace and can immediately begin creating the database schema.

It is worthy to note that the term database actually refers to a software application that can contain multiple tablespaces. Evolution Accounts each have at least one tablespace, but may share a database. The term database may be used in place of tablespace, because of its common understanding.

The database schema is the unique definition of an account's database structure. The schema is created via the myEvolution database tab using SQL statements.

The databases that Evolution Hosting offer fully support the SQL standard, in addition to offering proprietary extensions to do special tasks. Vendor specific documentation should be consulted for information on a particular vendor's SQL implementation.

Evolution Hosting recommends that account holders maintain a full set of SQL scripts (sometimes referred to as DDL) to create, recreate, or modify the account's database schema at any time. Ideally, these scripts should be maintained in a code versioning system, such as CVS.

Evolution Hosting provides tutorials, modifiable account properties, preconfigured application server setttings, and other tools to assist account holders in connecting to their database from Java applications. This section explains these tools in detail.

This section of myEvolution is where account databases can be modified, exported, and backed up.

Evolution Hosting offers full e-mail hosting facilities to account holders. These services can take the form of simple e-mail aliases, or dedicated e-mail servers with custom Servlet based management facilities. This section of the Evolution User Manual details the various options and their use.