THE BUSINESS ISSUETo reap all the benefits of automation, you also need to automate and secure deployment. All too often, companies look to other tools to cover deployment, overlooking the need for interoperability with Software Configuration Management.
THE ARCAD SOLUTION
With ARCAD, release management is just another logical, integrated part of the ALM process. Thanks to its highly integrated approach, the ARCAD-Skipper suite guarantees a strong bond and perfect coordination between development and production. With our solutions, you can: ARCAD coordinates deployment of native and non-native components in a single transfer. We can deploy programs and files in IBM i libraries at the same time as executables, DLLs, or any other files to any number of IBM i, Windows, UNIX, and Linux servers. Using our ANT interface, you can even write scripts to perform functions on target servers − start and stop a server to install an application, for example. Installation on remote servers automatically creates subdirectories identical to those in the reference environment.
With these functions, we cover both the needs of software vendors (patch and cumulative releases) and user companies that have a large number of machines (synchronized deployment).
ARCAD-Skipper aims for 100% automation. Procedures can be triggered at night and checked remotely. To guarantee security of automated deployment, ARCAD provides powerful rollback functions. This way, if a problem occurs during the installation of any of these components, the operation is cancelled, and our powerful integrated rollback functions return the application to a stable state as quickly as possible, whatever the host platform.
FEATURES
- Dispatch complete applications, cumulatives, or PTFs,
- Specific dispatch (list of objects or sources),
- Version level by site/application,
- Log of versions dispatched,
- Automatically confirm reception,
- Dispatch by TCP/IP, SNADS,
- Dispatch by Internet as emails,
- Dispatch in magnetic media format,
- Automatic reception handling on remote site,
- Log of actions on components,
- Log of executed commands,
- Error log with associated messages,
- Configure dispatches (time, jobq, etc.),
- Send/execute remote commands,
- Customize rollback processes,
- Multi-environment support on remote sites,
- Tracking tool for ongoing dispatches/receptions,
- Automatic rollback
HOW IT WORKS
To obtain 100% automation, ARCAD relies on its internal macro-language and execution agents. All default processes supplied with the product are easily customized, both on native IBM i and diverse platforms. An execution agent is a multi-platform processor used to execute remote commands that can be extended and integrated within a distributed business infrastructure. The agent allows implementation of new processes (or access to existing processes) by virtualizing a company’s multi-platform resources The execution agent must be able to provide services needed to manage primary tasks involved in the functioning of a remote process, and it must also take into account the different network infrastructures on which it will be executed (LAN WAN, VPN, etc.). It also needs to work within security constraints imposed by the receiving platform, and provide a configuration for integration with a distributed business network (access, LDAP, etc.). ARCAD Execution Agents
Our agents are developed on robust, open source “bricks” that are recognized for their reliability (DOM4J, Xerces, Log4j, ANT, and HSQLDB). They offer everything you need to ensure the system’s longevity and the reliability of future extensions. Authentication and authorization processes are delegated to customizable processes (by use of Java™ connectors) for integration into the business’s infrastructure design. 
Among other services, the execution agent provides file transfers, command execution, and SQL access.
- The file transfer service allows transportation of folders (or data) to and from the target machine (the one hosting the execution agent). For increased security, this service uses a proprietary protocol that is different from FTP.
- The execution service can handle execution of ANT scripts on the target system through secured, coherent communication links. It offers script transfers and execution in an embedded ANT interpreter, and sends back execution messages in real time. The agent authorizes definition of execution properties (ANT and JVM projects), access to these values, and access to external environment variables. It also offers the option to download ANT task or JAVA class libraries (to execute Taskdefs tasks), thereby increasing customization capabilities.
- SQL access services permit execution of SQL requests on JDBC-accessible databases on the domain server. It returns execution messages on performance of action requests (DDL, insert/update/delete, etc.) and executes selection requests with comeback information that includes data and metadata. The JDBC connection is customized by handling property passages and by loading JDBC pilots for access to the target database onto the server.
All logs are centralized on a single server to facilitate management and control. This log information is provided at a physical level (Which component has been installed? On which server? By whom? When? etc.), and also at the logical level (Which application version is installed on which site?). |
