Wednesday, June 27, 2012

Cell Broadband Engine

Mulch-Core Memory Coherence with bandwidth over latency is the main funda of CELL BROADBAND ENGINE ARCHITECTURE .It was introduced by Sony, Sony Computer Entertainment, Toshiba and IBM (STI) only for Network field, but now a days we find it in our very own Sony PS3 or pPlaystation 3.

Broadband Engine was first introduces with a vast 1 Tera- Flops power capacity within 90 nm Silicon on insulator process. Today's PlayStation 3 maintain 45 nm Cell processor architecture.It is as like 40 times faster than PlayStation 2.

In the PS3's Cell processor, there is one central core (the "Power Processing Element") while there are eight functional "synergistic processing elements". The Cell Broadband Engine is good at less generalized tasks than an X86 processor is but it is even worse at generalized tasks.

with the power of two-way simultaneous mulch-threaded main processor called the Power Processing Element (PPE) ,eight fully functional co-processors called the Synergistic Processing Elements(SPE) and a specialized high-bandwidth circular data bus

Some interesting features as fallow.........
    analysis of a broad range of workloads with cryptography.
   
    Graphical performance enhancement with lightning technology.

     fast-Fourier transforms (FFT).


OH yes......LINUX is here.IBM introduce and distributed the source code to support the hardware facility of CBE microprocessor with LINUX flavour.Not only that ,Sony Group also introduce gcc(GNU Compiler Collection)  and binutils for the SPU.The IBM Full System Simulator executable provides a rich set of capabilities for architecture simulation of the CBE microprocessor.Thread abstraction model on “Linux for CBEA" also can useble with the CBE SPE Management Library technology.


It is not only using in Sony's PLAYSTATION®3 but it already uses in Toshiba’s Cell Reference Set, a development tool for Cell products, and already is included in the IBM BladeCenter® QS20, or “Cell Blade,” as well as through joint collaboration with Mercury Computer Systems, Inc.It is also used as Video processing card, Supercomputing ,cluster computing, streem processing and distributing computing technology.

Saturday, June 23, 2012

RAID

Hello World,
   I am here again with a common topic RAID.I know it is a vast subject .I just try to disscuss it with a Nutshell.

So we all know RAID = Redundant array of independent disks or originally redundant array of inexpensive disks.What is this?
Actually It is a great way of storing data on multiple disk or in different place on network.Since multiple disks increases the mean time between failures (MTBF).

RAID technology was first defined by a group of computer scientists at the University of California at Berkeley in 1987.
Now a days Mainly 3 types of Array used in RAID technology.They are as follow.

Software based Raid
   Primarily used with entry-level servers, software-based arrays rely on a standard host adapter and execute all I/O commands and mathematically intensive RAID algorithms in the host server CPU.only used for low price and less component require.

Hardware based RAID
  Hardware-based RAID uses an intelligent drive controller and a redundant array of disk drives to protect against data loss in the event of media failure and to improve the performance of read/write operations. A disk array is an effective disk-storage solution for computers running database software.

External Hardware RAID Card
  Intelligent external array controllers "bridge" between one or more server I/O interfaces and single- or multiple-device channels.
 


There are mainly 2 types of RAID Level like Standard RAID and Nested or Hybrid RAID.Standard RAID include RAID 0, RAID 1, RAID 2, RAID 3, RAID 4, RAID 5, RAID 6 , RAID 7.

RAID 0
   Mechanism Used = Striping but no redundancy of data.
  
   Minimum Disk Required = 2

   Advantage = Improve performance and speed.

   Disadvantage = No fault tolerance.

RAID 1
   Mechanism Used = Mirroring or exact copy of date

   Minimum Disk Required = 2

   Advantage = Improve read performance and great fault tolerance for malty user system

   Disadvantage =Effective storage capacity become half and in software Raid we cannot use Hot swap disk management.

RAID 2
   Mechanism Used = Bit-level striping with dedicated Hamming-code parity or binary linear codes with error correction   parity.

   Minimum Disk Required = 3

   Advantage =It can recover from 1 drive failure or repair corrupt data or parity when a corrupted bit's corresponding data and parity are good

   Disadvantage =Entry level cost very high - requires very high transfer data rate.

RAID 3
   Mechanism Used = Byte level striping and dedicates one drive to storing parity information.

   Minimum Disk Required = 3

   Advantage = High throughput for large data transfer.

   Disadvantage =RAID-3 cannot overlap I/O. Very complex mechanism use and heavy resources used.

RAID 4
   Mechanism Used = Block level striping with parity.

   Minimum Disk Required = 3

   Advantage =The cost per megabyte of a RAID 4 array is low because only one drive in the array stores redundant data and     read speed is good.

   Disadvantage =Does not support multiple simultaneous write operations.


RAID 5
   Mechanism Used = Block level striping with distributed parity

   Minimum Disk Required = 3

   Advantage =Read data transactions are very fast but write data speed is quite slow

   Disadvantage =complex mechanism used as like RAID 3.

RAID 6
   Mechanism Used = This type is similar to RAID-5 but includes a second parity scheme.Used XOR parity generation and Read     Solomon ECC Generation.

   Minimum Disk Required = 4

   Advantage =extremely high data fault tolerance and can sustain multiple simultaneous drive failures.


   Disadvantage =It is quite expensive as a special controller is required for it.The dual parity checks increases the  complex calculations due to which the system performance

RAID 7
   RAID 7 is a Non Industry Standard proprietary product of a single company.

    It is mixed of a real-time embedded operating system as a controller with a caching via a high-speed       dedicated bus.The controller card uses its own CPU and combines striping and RAID-5-like storage.
  
  
Hybrid RAID or Nested RAID
   It is a combination of different standard level RAID used to increase high speed data throughput (Read and Write),disk   space manage and fault tolerance maintaining. They are also sometimes called two-dimensional.


RAID 0+1
  
   Mechanism Used = First the data are striped and next it mirrored.

   Minimum Disk Required =4

   Advantage = Higher performance than RAID 5 and easy installation

   Disadvantage =Inefficient of usable disk space and costly


RAID 1+0
  
   Mechanism Used = combination of RAID 1 and RAID 0. First Mirrored and Mirrored set distributed as block or striped.

   Minimum Disk Required = 4

   Advantage = High speed and good data security

   Disadvantage =costly and complex mechanism


RAID 50
  
   Mechanism Used = consist of series od RAID 5 and next striped.

   Minimum Disk Required = 6
  
  Advantage = It can sustain one to four drive failures while maintaining data integrity if each failed disk is in a different RAID 5 array.

   Disadvantage =very expensive


RAID 53
   Mechanism Used = Mirrored striped set with parity.
  
   Minimum Disk Required = 5
  
   Advantage = Very good I/O throughput for both data read and wright process

   Disadvantage =Expensive and critical mechanism implied


  There are so many Non-Standard level RAID :

•Double parity
•RAID DP
•RAID 1.5
•1.4 RAID 5E, RAID 5EE and RAID 6E
•1.5 RAID S or Parity RAID
•1.6 Intel Matrix RAID
•1.7 Linux MD RAID 10
•1.8 IBM ServeRAID 1E
•1.9 RAID-K
•1.10 RAID-Z
•1.11 Tahoe Distributed File System
•1.12 Drive Extender
•1.13 Drobo BeyondRAID


Fault Tolerance in RAID
  There are mainly 4 types of Fault Tolerance.They are as follow.......

    MTDL:
    Mean Time to Data Loss. The average time before the failure of an array component causes data to be lost or corrupted.

    MTDA:
    Mean Time between Data Access (or availability). The average time before non-redundant components fail, causing data        inaccessibility without loss or corruption.

    MTTR:
    Mean Time To Repair. The average time required to bring an array storage subsystem back to full fault tolerance.

    MTBF:
    Mean Time Between Failure. Used to measure computer component average reliability/life expectancy.


Mainly 3 types of Disk technology is used in RAID. They are as follow.

 Hot Plug Disk
 we can plug or unplug and Hard disk when the system is running. System can recognized and identify it.Mainly USB and IEEE  1394 can support it.

Hot swap Disk
  Hot swapping is used whenever it is desirable to change the configuration or repair a working system without interrupting its operation.we can apply Raid technology on it and repair.

Hot Spare Disk
  Hot spares are a good addition to a fault-tolerant array. They greatly reduce the impact of monitoring on the array reliability. when needed it added to the system to backup the server and when we repair the disk or change a new drive with a Hot plug or Hot swap technology it turns back to it idle state.




RAID technology apply




RAID 0 =RAID 0 is not really a "true" RAID system, as there is no fault tolerance involved here.
RAID 1 =Small Database system for Accounting service or Office uses.

RAID 5 =File and Application server, WWW, Intranet server
RAID 6 =Data Warehousing .

RAID 10 =High capacity disk, Internet server, SQL server

RAID 50 =File and Application server, Transaction processing, High capacity disk storage






Thursday, June 21, 2012

JINI...The big thing in Networking

JINI is a fantastic innovation for Network Industry.JINI is a great service oriented architecture with a great features like Scalability,Secure,evaluable and flexibility.It is a Sun Micro system innovation with a Java coding features similar to Java Remote Method Invocation.JINI allows for advanced searching for services through a process of discovery of published servicesJINI is a set of APIs and network protocols that can help you build and deploy distributed systems that are organized as federations of services.A service may be anything like Storage or Print.

Main Function of JINI......

A. It provides Share service to Users.

B. User can access resources anywhere within Network even the Network address may chainged.

C.Maintaining Services.

To enable JINI we need 3 items. Like JINI Service, JINI Locator and JINI client.......

JINI Networks work with the Unicast interaction or Dynamic Multicast discovery to find the service .It's also called LUS or Lookup service.