|
|
|
Referance
JiniSoft developed by Java RMI, JavaSpace & Jini Network Technology and become to MPP : Multiply Parallel Processing on Dynamic Distributed & Co-operative Systems. But current focus at Cross-Platform for Enterprise Application Server and developed by Java Platform, ISO C/C++, Objective-C, JavaScript & Microsoft .NET Famework ( including Windows RT )
4-Step mechanism for MPP : Multiply Parallel Processing
- Server Host must register to Engine Host ( LDAP : Active Directory Protocol ) and be informed to Engine Host at all times.
Dynamic Plug-in # Management will use information of Load Balancing to install the Business Logic or Service Factory inside Engine on any Server Host at low performance and suitable resource for each dynamic service, by Engine on Server Hoste be small static service for management.
- Service Proxy on Client Host must requested Server Address of Service Factory ( Dynamic Service ) from Engine Host through UDP, TCP or Web Service
- Engine Host ( cascaded structure with a DNS Server ) must sended Server Address and Digital Identity of Service Factory to Service Proxy under Transaction @ Request which is based on Load Balancing and Performance of each Server Host
Usually start @ Transaction Login, next End User choosed to Transaction Flow by their duties and responsibilities. Which defined in Human Relation & Human Resource Management System
So JiniSoft include OODBC Engine 2.0 : Object-Oriented DataBase / MediaBase / LogicBase, Distributed & Co-operative Computing, Enterprise Application Server and Natural Security on Human Relation together until the Application.
- Communication with Client Host and Server Host through Service Proxy & Factory by Remote Logic Interface & Dynamic Object Serialization on TCP/IP. Regardless with Business Rules, Transaction Flow or Process
Differences Developement.
- Develop by Java Paltform will use Java RMI, JavaSapce or Jini Network Technology, depending on the difficulty of installation, performance, applications and environmental ( of quality, monitoring, mechanism ) on the Network.
- Develop by Microsoft .NET and Windows RT use Binary .NET Remoting on TCP/IP.
- Develop by Native ISO C/C++ or Objective-C use Socket Protocol through Serialization with Big-Endian to communicate with Java Platofrm & ARM/RISC Architecture and Little-endian to communicate with Microsoft Platform and x86/CISC Architecture.
- For Customize and Interface with other Application use Web Service ( SOAP / REST / JSON ), but it is slower than Binary Protocol more than 10 times.
Record Cache in Service Proxy available in 3 forms.
- For Desktop, It is only Cache Memory in client application
- For iPad or Tablet, It will save on Flash Storage for Backup & Mirror with Compress to support large scale of Information, Object or Record, and Cryptography to confirm the property and creators in every Record collectively called JiniSoft File Systems
- For Proxy Server to Backup & Mirror and service to Client Host or all desktop on network to reduce the data transfer. Or move service ( Transactions & Tables ) all or part of the local & combined with branch data for Private Cloud Computing.
Vision of Transaction Flow ( in Service Proxy & Service Factory )
Query is the mechanism of Business Rules in Object-Oriented SQLs together until Transaction, and each Transaction as defined under Object-Oriented Workflow called Transactions
Result of Query is Arrary Objects inside OODB Engine and next to unlimited Query. The key mechanism is Cache Management as Virtual Memory of OS : Operating Systems, called Virtural Table by Swap the Object or Record @ like or often from storage to Memory and each Table of DataBase endlessly divided into Sub-section or multiply section, they called Multiply Cache
Multiply Cache endlessly process the Virtual Style on TCP/IP, become to MPP : Multiply Parallel Processing on Dynamic Distributed & Co-operative DataBase
Single Core for IaaS : Infrastructure as a Service, PaaS : Platform as a Service & SaaS : Software as a Service ( Approximation @ 10 MB )
Difference between JiniSoft with Compiere and Apache OFBiz
For Compiere developed by the former developer of SAP, or ERP \ CRM Open Soruce developed on Apache OFBiz, they developed by Java Platform and Business Rule on client through JDBC : Java Database Connectivity by Shared Database, if clients have large number, ability is limited by performance depended spec of hardware together client, server and capacity of network.
Currently, developement of hardware is limited.
Remark Multi-Core Processor not help softwrae to faster and bottleneck @ Data Buse of RAM & Storage, Exception divied into sub-data to process in cache memory of processor.
|
Last update : March 09, 2013 12:00 ( Thailand )
Java and all Java-based are trademarks of Oracle Corporation.
Apple, Mac, iMac, iPhone, iPod and iPad are trademarks of Apple, Inc.
Windows, .NET or all Microsoft-based are trademarks of Microsoft Corporation.
JiniSoft Corporation Limited
Copyright @ 1990 - 2013 Mr. Roongroj Rojanapo (  )
18/81 Moo 9, Bang Pra, Si Racha District, Chon buri 20110, Thailand
|
