Figure 1 binding and bonding the internet
The internet, as I see it, is: the world’s largest defined digital communications framework. It initiates with protocols, these develop into standards, and this facilitates a rapid [communications process]. So, as a start point we have our data to which the first element we are going to need from the framework is a wrapper in the form of TCP from the protocol stack suite. After that: we are going to bind IP; from this point forwards it is routing and the addition of computational mathematics and hay presto.
So from our diagram, we have, a framework called the internet, within in which are other frameworks; some of the frameworks are stand alone, others interact and are notionally bound. [layering]
Figure 2 the convergence of framework ideas
As an understanding: grid computing is: the application of multiple networked computational resources to simultaneous problem solving, often involving a single problem domain.
The process requires management software that is able to divide and farm out elements of the programme to the farm. It can be considered as: distributed large-scale cluster computing as forms of network distributed parallel and multi-parallel processing.
Therefore, at its simplest grid computing is a network of computing resources within which each resource is shared with all resources.
Grid computing stems from distributed computing.
An ideal grid would be one where the networked resource represents one computer.
Most grid systems rely on proprietary software and tools.
· central processing unit
· network access
It boils down to basic programming, together with the design of arrays to provide resolution. For example the Allen array.
SETIMessage = "00000010101010000000000 00101000001010000000100 10001000100010010110010 10101010101010100100100 00000000000000000000000 00000000000011000000000 00000000001101000000000 00000000001101000000000 00000000010101000000000 00000000011111000000000 00000000000000000000000 11000011100011000011000 10000000000000110010000 11010001100011000011010 11111011111011111011111 00000000000000000000000 00010000000000000000010 00000000000000000000000 00001000000000000000001 11111000000000000011111 00000000000000000000000 11000011000011100011000 10000000100000000010000 11010000110001110011010 11111011111011111011111 00000000000000000000000 00010000001100000000010 00000000001100000000000 00001000001100000000001 11111000001100000011111 00000000001100000000000 00100000000100000000100 00010000001100000001000 00001100001100000010000 00000011000100001100000 00000000001100110000000 00000011000100001100000 00001100001100000010000 00010000001000000001000 00100000001100000000100 01000000001100000000100 01000000000100000001000 00100000001000000010000 00010000000000001100000 00001100000000110000000 00100011101011000000000 00100000001000000000000 00100000111110000000000 00100001011101001011011 00000010011100100111111 10111000011100000110111 00000000010100000111011 00100000010100000111111 00100000010100000110000 00100000110110000000000 00000000000000000000000 00111000001000000000000 00111010100010101010101 00111000000000101010100 00000000000000101000000 00000000111110000000000 00000011111111100000000 00001110000000111000000 00011000000000001100000 00110100000000010110000 01100110000000110011000 01000101000001010001000 01000100100010010001000 00000100010100010000000 00000100001000010000000 00000100000000010000000 00000001001010000000000 01111001111101001111000"
I have a way to communicate with the computer: “grid commuting” becomes: 0110011101110010011010010110010000100000011000110110111101101101011100000111010101110100011010010110111001100111
SETI sent the above message from the Arecibo Observatory and the 16th of November 1974, it was transmitted in the direction of:
Where a little bit away and happens to be Messier object, the globular star cluster M13 [NGC 6205], which according to best estimates is about 6,800 parsec away, in the constellation Hercules.
Figure 3 Messier Object M13
It has an apparent dimension 20 arcmin, with an apparent magnitude of +5.8
Mass is estimated at 6x105 solar masses, it has a radius of 25.76 parsec with an age of 11.65x109 years
Figure 4 Constellation Hercules
The constellation has an area 1225 square degrees, and there are nine stars within 10 pc, it was one of the 48 constellations listened by listed by Ptolemy, and remains as one of the current “modern” 88 constellations
The history of patterns in the sky that we as constellations, has been one of enduring legacies with these ideas passed generation to generation civilisation to civilisation and then on through history to the generations that follow.
This modern depiction of Hercules, is an amalgamation of previously described constellations, the Babylonians had ideas and descriptions, described technically as: conflation: for the constellation. These constellations have been developed through time as objects in space and in our history through mankind’s most ancient civilisations. Amongst the first of these the Sumerians, had their own extra ordinary and detailed descriptions, ideas, mythologies for the region in space that we describe as the constellation Hercules.
Figure 5 a bit of rewind and fast forward
So we get to:
./<index>0« to ︣» and apply this notational expression to grid computing algorithm.
So if we ask Charles Babbage, who originally created a difference engine.
So what I understand is the Internet, comprises:
This is a set of communications protocols used for the Internet and similar networks; the most common stack of which is TCP/IP. It provides: end to end connectivity. So we have:
Figure 6 point to point
It is used in specifying how data should be formatted together with addressing: transmitted, rooted and received at the destination.
It is abstracted into four layers:
This leads us to the OSI model:
6. Data link
The addition of the physical layer.
The analysis of huge datasets together with the ability to run scenario analysis at unprecedented rates, growing both their range; increasing granularity, resolution in the focus in scope increasing their depth exponentially growing rate. ()
That detailed results require today.
· UNIX servers
· Intel servers
· Storage systems
· Desktop PCs
Fundamentally this list is extensible into any computing device with network access.
Figure 7 unused processor cycles non-grid
Others are hybrid. This provides the opportunity for the creation of grid middleware.
Figure 9 grid computing
Most of the time computers have a lot of available resources.
Note: the computational resources do not need to share the same physical location.
Lots of “little” computers used in simultaneous arrays turn into “super” computers.
Grid is different from cloud.
A grid definition: co-ordinated resource sharing for problem domain solving in dynamic, muti-institutional virtual organisations.
We just want to solve a problem.
· Distributed system
· Site autonomy
· Systems management
· Resource management
· Data management
· Information services
Pool of computational resources and a “seeker” or user, wanting to solve a problem.
Time and the application of unused resources.
Figure 10 CERN representation
CERN is made up of multiple grids, the facility in Geneva, Switzerland is currently configured to provide about 20% of grid functions for the analysis of LHC data, the remainder is provided by:
· organisational contributors
*instruments within the infrastructure
It’s working now!
The model basis is: national infrastructure supported by regional collaboration.
Although the system is up and running there are a few limiting elements, these comprise of:
· need fast interconnection between computational resources
· administrative domains
· politics of sharing
An attack is equal to an assault on system deliberately avoiding security systems.
Security in the computational science is principally information security. Computer security also includes the fields of contingency planning and disaster recovery.
So from the ground up we may start with a design plan, fortunately for us, there is a recognised structural template in place, and it includes the following techniques:
· the principle of least privilege
· automated theorem proving
· code reviews and unit testing
· defence in depth
· failsafe and full-back positioning
· audit trails
· window of vulnerability minimisation, through full disclosure
This now implies that within our information technology architecture we now have a security layer the artefacts of which describe management of the systems quality attributes.
· insurance services
Figure 11 grid security issues
A denial-of-service attack (DoS attack) or distributed denial-of-service attack (DDoS attack) is an attempt to make a machine or network resource unavailable to its intended users.
· ICMP Flood
· (S)SYN flood
· low rate denial of service
· asymmetry of resource utilisation in starvation
· permanent denial of service
· application level floods
· HTTP post denial of service
· slow read
· reflective or spoofed
· telephony denial of service
· unintentional denial of service
· denial of service level II
· application front end hardware
· IPS-based prevention
· DDS-based defence
· blackholing and sinkholing
· clean pipes
Figure 12 Example of a defensive system for denial of service attack