fabric server

FabricServer Technology Overview - Presentation Transcript

1. Package, Deploy and Manage Enterprise Applications on Physical, Virtual or Cloud-Based Environments FabricServer
2. FabricServer
   • FabricServer is management software used to package, deploy and manage enterprise applications on physical, virtual or cloud based datacenter environments
   • FabricServer extends the principles of DataSynapse’s market leading Grid computing platform:
   • Shared and dynamic infrastructure
   • Automated application packaging and deployment
   • Built in application resiliency and failover
   • And adds support for virtualization, utility and cloud computing for enterprise applications based on standard development platforms (e.g. JEE, Microsoft apps, etc..)
   • Benefits of using FabricServer:
   • Lowering application time to deploy for complex enterprise applications
   • Lowering infrastructure costs, through improved utilization of data center resources
   • Improving application performance against SLAs
   • Driving adoption of virtualization (e.g VMware) and external cloud models (e.g. EC2)
3. FabricServer - Basics FabricServer Package Once FabricServer provides tools and templates required to build self-contained, runnable packages of applications and platforms Tools Self-Contained Packages Configuration Information Platforms Packaged Legacy FabricServer Run Anywhere FabricServer manages those packages and can activate on any runtime environment (physical, virtual, cloud) Clouds Virtual Physical Automated Deployment Service Repository Run-Time Broker Policy Engine Manage Runtime FabricServer allows production support teams to create policies for how applications will be managed at run time Closed Loop Policy-Based Allocation of Infrastructure
4. Run-Time Architecture FabricServer Broker Failover Fabric Broker Shared Pool of FabricServer Hosts Daemon
   • Monitors host & spawns one or more engines
   • 1 per physical host)
   Engine
   • Manages application instances
   • Isolates applications on hosts
   • 1 or many per physical host
   Component
   • A application component instance
   • Multiple components = application
   • 1 per engine
   <> Engines are only started as needed VM VM VM Engines can be deployed to physical or virtual servers The entire stack of an enterprise app will typically be distributed across different engines
5. Brokers and Engines
   • 100% server based Java Application
   • Runs on Windows, Solaris, HPUX, AIX and most Linux distros
   • Requires a standard J2SE
   DataSynapse Engine
   • Runs on Windows, Solaris, and most Linux distros
   • Started and managed by an agent that runs on the host called the Engine Daemon.
   • Daemon starts and stops engine instances based on Policy
   • Engine Daemon starts Engine instances, monitors them and restarts them if there are failures or reconfigurations
   • DataSynapse Engine:
   • Starts/stops/monitors applications in other processes
   • More than one engine instance can run per host
   Operations Dashboard Application Component and Library (with SDK) Policy Manager Auditing and Logging Allocation Engine Statistics Manager Real-time Systems Monitor Security Manager FabricServer Broker Host Engine Daemon Engine Instance <>
6. FabricServer Tools
   • DataSynapse Studio:
   • An eclipse based UI for infrastructure architects to design and build application infrastructure templates
   • Key Features:
   • Graphical and command line tools simplify creating and updating FabricServer artifacts
   • Wizards and structured XML file editing ease the creation of components and configuration templates
   • Libraries can be easily compared and modified
   • Applications can be smoke tested within Studio before deploying to a full FabricServer environment
   • FabricServer UI:
   • A web application UI for production support personnel to manage and monitor a FabricServer environment
   • Key Features:
   • Create and activate policies and schedules
   • Monitor and manage all applications running on FabricServer
   • Manage utility environment and configuration
   • Diagnose, log, and configure complex systems
7. Pre-Built Artifacts
   • Containers:
   • The controlling logic for the application server environment that configures, starts, monitors, and shuts down the processes.
   • SDK that enables you to create your own containers
   • Distributions:
   • The application component run-time “Golden Image” packaged as a single image
   • These images contain no DataSynapse code or modifications to the vendor package
   FabricServer Repository Distributions Containers Components … … SDK …
   • Components:
   • The app deployment artifacts, platform configuration, and runtime context required to run anywhere in a FabricServer environment
   • Applications:
   • Collection of components that make up an application, e.g. WAS Deployment manager, Cluster of managed servers, MQSeries
8. Supported Containers
   • Supported Containers
   • Full production support for existing releases
   • New releases are shipped with core DataSynapse product release
   • Updates to ISV application are supported within 9 months
   • Licensed and charge as per standard price list
   • Shared Source Containers
   • “ Starting point” for developing a custom container
   • Source code shared in an online code repository
   • Updates and maintenance provided on SOW basis
   • No software license, no software license charge

Labels: