Tumblrs for Humanity Unite!
Most fans of science and academia would agree that Berkley University is one of the top institutions when it comes to research.
One of their defining achievements is BOINC (Berkeley Open Infrastructure for Network Computing). For those unfamiliar, BOINC is a Distributed Computing Platform, which uses distributed systems to solve computational problems. In distributed computing, a problem is divided into many tasks, each of which is solved by one computer. Essentially many computers around the internet can work together to solve small pieces of a much bigger problem.
Researchers found that distributed computing can be used to solve all kinds of problems, including those found in mathematics, medicine, molecular biology, climatology, and astrophysics; and volunteer computing was born.
Computer users on various platforms all over the world donate spare processing power from their computers, when idle. If you are surfing online, working on a project, or playing a game; the distributed computing software stays dormant, and does not affect your user experience. Volunteer computing is like donating to a worthy cause without any cost to you.
Today, we created a new account with World Community Grid, which is hosted on BOINC, and hosts some great causes like Computing for Clean Water; The Clean Energy Project; Help Cure Muscular Dystrophy; Help Fight Childhood Cancer; Help Conquer Cancer; Human Proteome Folding; and FightAIDS@Home. Help Fight Childhood Cancer is a cause near and dear to my heart, but there might be a cause that appeals more to you.
I encourage you to take action. Download and install secure, free software that captures your computer’s spare power when it is on, but idle. You will then be a World Community Grid volunteer. Use your Tumblr screen name if you want, and help promote your blog.
We also created a team, called Tumblrs for Humanity. Please join so we can see the good work that socially minded bloggers can do!
My hope is that you will not only join this great cause, but that you will spread the word. Would you kindly consider joining and/or re-blogging this to spread the word to the Tumblr community?
450 000 man-machine artists in collaboration
An article by @ilamandarina on participation in science made me think that this artist, Scott Draves, was working along similar lines (from here).
Scott Draves creates art by writing software that runs an internet distributed supercomputer consisting of 450,000 computers and people, creating images as a form of artificial life, each with its own genome, generated by thousands of numbers that define how it looks and moves. The first versions of this algorithm date from 1992.’
[…]
First created in 1999 by Scott Draves, the Electric Sheep is a form of artificial life, which is to say it is software that recreates the biological phenomena of evolution and reproduction though mathematics. The system is made up of man and machine, a cyborg mind with 450,000 participant computers and people all over the Internet.
This is a distributed system, with all participating computers working together to form a supercomputer that renders animations, called “sheep”, that everyone sees. The human participants guide the survival of the fittest by voting for their favorite animations in the flock. You can join this project by downloading the Electric Sheep Screensaver.
Clustering RabbitMQ
Clustering RabbitMQ is very easy - if you know how. Unfortunately, the documentation on this topic is good but not good enough (cf. RabbitMQ Clustering). If you try to do it, you may get lost on the track until you find some insightful posts on the mailing list. This is why I summarize here how I got it to work.
Say, you want to create a cluster having two disc nodes and two ram nodes. If you do this on at least two machines, each having a disc and a ram node you achieve good fault tolerance and good scalability both with one setup. Your clients may connect to the ram nodes only or these are balanced by an additional load balancer.
But, how do I make a node a disc node and another node a ram node?
There’s no such command like “rabbitmqctl mkdisc” and there is no related configuration option. On one hand, this is a little counter intuitive, on the other hand this adds a lot of flexibility since you may alter the roles of nodes and restructure your cluster on the fly whenever necessary.
The rules are assigned by the way you call the “rabbitmqctl cluster” command. In our scenario, we have multiple nodes on the same host, so we need to wrap the calls to “rabbitmqctl” into shellscripts setting some environment variables (cf. RabbitMQ Configuration). If this has been done, you ensure all nodes of the cluster are running. Afterwards you execute a sequence of “stop_app”, “reset”, “cluster”, “start_app” commands for all nodes. If it comes to the “cluster” command, you add a space separated list of all disc nodes you want to create to the “cluster” command executed for each node. My mnemonic for this is that you copy the current node to all disc nodes. The whole sequence may look like this, with “rbctl.*” being your wrapper scripts:
host-of-disc1$ rbctl.disc1 stop_app
host-of-disc1$ rbctl.disc1 reset
host-of-dics1$ rbctl.disc1 cluster disc1@host-of-disc1 disc2@host-of-disc2
host-of-disc1$ rbctl.disc1 start_app
host-of-ram1$ rbctl.ram1 stop_app
host-of-ram1$ rbctl.ram1 reset
host-of-ram1$ rbctl.ram1 cluster disc1@host-of-disc1 disc2@host-of-disc2
host-of-ram1$ rbctl.ram1 start_app
host-of-ram2$ rbctl.ram1 stop_app
host-of-ram2$ rbctl.ram1 reset
host-of-ram2$ rbctl.ram1 cluster disc1@host-of-disc1 disc2@host-of-disc2
host-of-ram2$ rbctl.ram1 start_app
host-of-disc2$ rbctl.disc2 stop_app
host-of-disc2$ rbctl.disc2 reset
host-of-disc2$ rbctl.disc2 cluster disc1@host-of-disc1 disc2@host-of-disc2
host-of-disc2$ rbctl.disc2 start_app
If you have to add users, vhost and permissions, you better do it at the end of this procedure, otherwise the “reset” will delete all of this information. Also, if you want to change the cluster setup later, you should be careful with “reset”, omitting it for one disc node at least.
Another weak point with the whole clustering stuff is the location of the “.erlang.cookie” file. This file is essential for clustering and must have the same content for all nodes in the cluster. Documentation says RabbitMQ looks at “/var/lib/rabbitmq/.erlang.cookie” but I found this not always true. Supposed RABBIT_HOME points to the directory where the rabbit distribution is located, I copied the file to “$RABBIT_HOME/../.erlang.cookie” and RabbitMQ used this one. I’m not quite sure if this is a general rule.
On Collective Action, Distributed Computing, and Volunteerism: A Praise of Distributed Computing Projects
Your computer that you are using is a most ingenious and powerful device. With it, you have access to virtually unlimited amounts of information, of computational capability, and a variety of other things. Yes, you are fortunate, oh user of the Internet. Never before in history have entire worlds been at the tips of one’s fingers.
With all the power and inter-connectivity of the internet, it was only a matter of time before some intelligent individuals realized that they could use the internet to aid in massive, humanitarian projects, ones that benefit all of mankind. Their idea was simple: One personal computer by itself cannot process all the research and such required to undertake these projects. But what if you took these projects, and broke them down into very small pieces, and then had many computers work on these individual pieces?
In 1996, this concept, called “distributed computing”, first took the form of the Great Internet Mersenne Prime Search. Due to the infinite nature of numbers, this project continues to this day. Computers around the world are contributing bit by bit to finding the next Mersenne Prime, and the results have been tremendous. Mersenne Primes 29 through 40 have been discovered via this project, something that would take normal research much, much longer to discover. The real beauty of this project is that those who are letting their computers be used are doing so voluntarily. No financial compensation is expected, rather, they are willingly donating their computational power to a great problem.
Of course, a search for large numbers, while interesting and possibly useful, can only do so much for humanity. But the thing is, this concept of “distributed computing” has caught on. Now, we have real, substantial projects with goals of fighting disease and poverty around the world. The most notable of these projects is the World Community Grid. This project, started by IBM, has had an enormous, positive impact on research into disease control and treatment, water treatment, clean energy, and more. When the project was started, it was aimed at simply finding a cure for smallpox, in case that disease ever returned. Using distributed computing, 35 million potential cures were analyzed over a period of time. In the first 72 hours, 100,000 results were returned. The downtime of about 2 million computers was involved, and resulted in scientists finding 44 potential strong candidates. But why is such a project so ingenious?
The biggest thing is costs. It normally takes millions upon millions of dollars to find drug compounds that are effective against a disease. This concept spreads those costs out over millions of individuals, who only suffer a slight increase in utility usage as a result. These programs are designed to stop using a computer whenever it is overheated, or is being used for pretty much anything else. As I type this, my computer is being used in a small way for one of the many World Community Grid projects. Whenever my usage of the computer reaches a certain percentage of the CPU, the project suspends usage of my computer until it goes back below that level. As a result, it costs far less to discover new solutions to major problems, which benefits everyone.
The next biggest thing is how this solves collective action problems. Normally, it is difficult to get people involved in such projects outside of a computer, due to the cost to their time and energy. But with distributed computing, you increased the efficacy of each individual’s contribution (your computer might be the one that discovers the cure for HIV/AIDS) while keeping personal costs extremely low. This helps explain why so many individuals are helping contribute to help fight problems they normally would be unlikely to help with.
The way in which you can contribute is free. All you need to do is download one of the distributed computer programs that have been created. The most common and popular program is BOINC. This program will also allow you to search out specific projects you wish to work on. These projects include: SETI@home, where you analyze data received by the SETI program to help find extraterrestrial intelligence; the World Community Grid, which I have explained; Cosmology@home, where your computer helps analyze models of the universe to find the best one, and many many more.
You have been fortunate in life to have such a powerful device as a computer. This program doesn’t require much of you or your computer, and it actively helps great causes. Why not give it a try?
Netflix @ DataStax SF 2011 - Monday, July 11
Netflix will be sponsoring DataStax SF 2011 on Monday, July 11. Though Netflix does not often sponsor events or recruit at conferences, we hope to engage with others active in the Cassandra and Distributed Systems community.
Stop by our booth to learn about what we are working on! Bring your resume as well!
Sid, Cloud Systems, Netflix
Books on Science - Essays Inspired by Microsoft’s Jim Gray, Who Saw Science Paradigm Shift - NYTimes.com
nytimes.com“Dr. Wing has argued that ideas like recursion, parallelism and abstraction taken from computer science will redefine modern science. Implicit in the idea of a fourth paradigm is the ability, and the need, to share data. In sciences like physics and astronomy, the instruments are so expensive that data must be shared. Now the data explosion and the falling cost of computing and communications are creating pressure to share all scientific data.
“To explain the trends that you are seeing, you can’t just work on your own patch,” said Daron Green, director of external research for Microsoft Research. “I’ve got to do things I’ve never done before: I’ve got to share my data.””
LHC@home 2.0
boinc01.cern.chMake your mark in science, by utilizing your graphics card during idle times to simulate the Large Hadron Collider through distributed computing.
Similar projects include Folding@home (protein folding) and SETI@home (extraterrestrial intelligence research).
Birazcık scaling: Python ve Django uygulamalarında asenkron işler (RabbitMQ)
Konuya girmeden önce; artık blog post’larım django.org.tr gezegeninde yayınlanmakta. Bu yüzden bana tumblr’ın getirtiği baştan salma yazı yazma alışkanlığımı sonlandırmak istiyorum. Başlıksız ya da içeriksiz post atmak her ne kadar güzel olsa da post’lar sadece tumblr üzerinden okunmuyor. Artık özen göstererek yazmak gerekecek :)
Distributed Computing
RabbitMQ, zaten dağıtık hesaplama (distributed computing) konusunda ünlü olan Erlang dili ile geliştirilmiş AMQP(Advanced Message Queuing Protocol) protokolü üzerinde çalışan open-source bir message broker (mesaj kuyruğu diyebiliriz) yazılımıdır. Bu ve bunun gibi message broker yazılımları sayesinde kullandığımız dilden bağımsız bir şekilde yapılması uzun süren işlemleri (hesaplamalar, email vb.) çeşitli makinalara ya da aynı makina üzerindeki birden fazla worker’a dağıtabiliriz.
Gerçek hayattan bir örnek verecek olursak; scale edilmesi gereken bir django uygulamasında request ve response arasında uzun süren işlemler ya da web ile alakası olmayan ayrı bir katmanda yapılan bir iş olmaması gerekmektedir. Misal kullanıcılara bir ya da birden fazla email göndermek, uzun süren hesaplamalar yapmak, kullanıcıya yeni bir demo site açmak gibi …
Requirements;
- Erlang platformu (Linux üzerindeyken RabbitMQ kendisi yüklemekte. Eğer windows üzerinde iseniz şuradan indirip kurabilirsiniz.)
- RabbitMQ Server (Windows’ta iseniz şuradan indirebilirsiniz)
- Kullanacağımız programlama dili için RabbitMQ client kütüphanesi. Örneklerde python için pika kullanacağız. Diğer diller için kütüphaneler şurada.
Paket manager’ımız ile rabbitmq-server’ı yükleyelim.
apt-get install rabbitmq-server
Kurulum bittiğinde server otomatik olarak başlatılacaktır. Şimdi de python kütüphanemizi yükleyelim.
pip install pika
Bu işlem de bittikten sonra yükleyeceğimiz başka bir şey kalmıyor. Örneklere geçebiliriz.
Bir mesaj kuyruğu uygulamasında üç temel yapı vardır; Consumer, Queue ve Publisher.
Consumer
Publisher’dan gelecek olan mesajlar için sürekli dinlemede olan bir nevi sunucudur. Aynı anda birden fazla consumer çalıştırabilirsiniz. RabbitMQ tüm consumer’lara eşit miktarda iş yükü dağıtmaya çalışacaktır.
Queue
Task yani görevlerin saklandığı kuyruktur. Eğer tek consumer var ise bu yapı FIFO (First in First out) yani -ilk giren ilk çıkar- şeklindedir. Zira birden fazla consumer olduğunda RabbitMQ işleri dağıtmaktadır.
Publisher
Consumer’a işleri gönderen uygulamadır. Bu bir django application’ı olabilir.
Şimdi örneğimize geçelim; peş peşe email gönderen bir uygulama simüle edelim. İlk olarak receiver yani consumer uygulamamızı yazalım.
consumer.py içeriği;
import pika, time
def callback(ch, method, properties, body):
print "email gonderiliyor; ", body
time.sleep(1) # gercekci olsun diye 1 saniye bekletiyoruz :)
print "email gonderildi."
def main():
connection = pika.BlockingConnection(pika.ConnectionParameters(host='localhost'))
channel = connection.channel()
channel.queue_declare(queue='mailing')
channel.basic_consume(callback, queue='mailing', no_ack=True)
channel.start_consuming()
if __name__ == "__main__":
main()
Örnekte mailing adında bir kuyruk oluşturduk ve o kuyruk üzerinde consuming işlemini başlattık. Örneği çalıştırdığınızda program dinlemeye geçecektir. Bir mesaj geldiğinde ise oluşturduğumuz callback fonksiyonu mesaj parametresi ile birlikte işlemeye başlayacaktır. Consumer uygulamamızı aşağıdaki gibi başlatalım;
python consumer.py
Şimdi ise publisher’ımızı yazalım.
publisher.py içeriği;
import sys
import pika
def main():
connection = pika.BlockingConnection(pika.ConnectionParameters(host='localhost'))
channel = connection.channel()
channel.queue_declare(queue='mailing')
channel.basic_publish(exchange='', routing_key='mailing',body=sys.argv[1])
connection.close()
if __name__ == "__main__":
main()Consumer uygulamamızdaki gibi aynı kuyruğa bağlandık ve bu sefer kuyruğa mesaj göndermesini istedik. Publisher uygulamamızı aşağıdaki gibi çalıştırdığımızda dinlemede olan consumer’ımız üzerine görevi yapacaktır.
python publisher.py fatiherikli@gmail.com
Burada önemli olan kısım publisher’dan ziyade consumer uygulamamızın yapılandırmasıdır. Örnekte biz tek bir consumer uygulaması çalıştırdık. Peş peşe publisher.py üzerinden komut göndermeye çalıştığınızda Consumer’ınız sırasıyla gönderdiğiniz komutları çalıştıracaktır.
Şimdi ise consumer.py ‘nizi farklı terminallerde aynı anda çalıştırın. Örnek olarak 3 tane consumer’ınız dinlemede olsun. Publisher’ınız üzerinden peş peşe mesaj göndermeye çalıştığınızda her consumer’da aynı görev sayısı olacak şekilde işlemlerin dağıtıldığını göreceksiniz.
Ayrıca Django ve RabbitMQ etkileşimini daha da kolaylaştıran celery projesini incelemenizi tavsiye ederim;
Network Architecture - What is Network Architecture?
Network architecture design.Network architecture is the design of a communications network. The network architecture of the Internet is predominantly expressed by its use of the Internet Protocol Suite, rather than a specific model for interconnecting networks or nodes in the network, or the usage of specific types of hardware links.
OSI Network Model
1.1 Physical Layer
1.2 Data Linking Layer
1.3 Network Layer
1.4 Transport Layer
1.5 The Session Layer
1.6 Presentation Layer
1.7 Application Layer
OSI Network Model
On each layer, an instance provides services to the instances at the layer above and requests service from the layer below.
Physical Layer
The Physical Layer defines the electrical and physicalspecifications for devices. This includes the layout of pins, voltages, cable specifications, hubs, repeaters, network architecture adapters, host bus adapters (HBA used in storage area networks) and more.
Data Link Layer
The Data Link Layer provides the functional and procedural means to transfer data between network architecture entities and to detect and possibly correct errors that may occur in the Physical Layer. Network architecture Layer
The Network Layer provides the functional and procedural means of transferring variable length data sequences from a source host on one network to a destination host on a different network, while maintaining the quality of service requested by the Transport Layer (in contrast to the data link layer which connects hosts within the same network). The Network Layer performs network routing functions, and might also perform fragmentation and reassembly, and report delivery errors. Routers operate at this layer sending data throughout the extended network architecture and making the Internet possible. Transport Layer
The Transport Layer provides transparent transfer of data between end users, providing reliable data transfer services to the upper layers. The Transport Layer controls the reliability of a given link through flow control, segmentation/desegmentation, and error control. The Session Layer
Presentation Layer
The Presentation Layer establishes context between Application Layer entities, in which the higher-layer entities may use different syntax and semantics if the presentation service provides a mapping between them. If a mapping is available, presentation service data units are encapsulated into session protocol data units, and passed down the stack.This layer provides independence from data representation (e.g., encryption) by translating between application and network architecture formats. The presentation layer transforms data into the form that the application accepts. This layer formats and encrypts data to be sent across a network architecture. It is sometimes called the syntax layer. Application Layer
The Application Layer is the OSI layer closest to the end user, which means that both the OSI application layer and the user interact directly with the software application. This layer interacts with software applications that implement a communicating component. Application layer functions typically include identifying communication partners, determining resource availability, and synchronizing communication. When identifying communication partners, the application layer determines the identity and availability of communication partners for an application with data to transmit.
Distributed computing
In distinct usage in distributed computing, the term network architecture often describes the structure and classification of a distributed application network architecture, as the participating nodes in a distributed application are often referred to as a network architecture. For example, the applications network architecture of the public switched telephone network (PSTN) has been termed the Advanced Intelligent Network. P2P networks usually implement overlay networks running over an underlying physical or logical network architecture. These overlay network may implement certain organizational structures of the nodes according to several distinct models, the network architecture of the system.
Autonomic Network Architecture.The Autonomic Network Architecture Project aims at exploring novel ways of organizing and using network architecture beyond legacy Internet technology. The ultimate goal is to design and develop a novel autonomic network architecture that enables flexible, dynamic, and fully autonomous formation of network nodes as well as whole network architecture. The resulting autonomic network architecture will allow dynamic adaptation and re-organization of the network according to the working, economical and social needs of the users. The ultimate goal is to design and develop a novel network architecture that enables flexible, dynamic, and fully autonomic formation of network nodes as well as whole network architecture. The scientific objective of this proposal is to identify fundamental autonomic network architecture principles. Moreover, this project will build, demonstrate, and test such an autonomic network architecture. Project Objectives
Scientific Objective
To identify fundamental autonomic networking principles that enable network architecture to scale not only in size but also in functionality. A new Autonomic Network Architecture will emerge as a result of this research. Technological Objective
The technological objective of ANA is therefore to build an experimental autonomic network architecture, and to demonstrate the feasibility of autonomic networking within the coming 4 years.
The goal is to demonstrate self-organization of individual nodes into a network architecture. The design of such network architecture should potentially scale to large network meshes in the range of 105 active (routing) elements. Here the focus is on the self-organization of network architecture into a global network.
Global Information Network Architecture.The Global Information Network Architecture (GINA) Team was created in 2004 to address this possibility. The Global Information Grid
Vector Relational Data Modeling (VRDM) Modeling Models.
The GINA Team created GINA to enable model-based software engineering, but we did so in such a way that the model, once-defined, represented a working application. Moreover, the GINA team made the decision early in its development to make GINA a GINA model. Enabling GINA’s deep configurability required the development and implementation of multiple models.
The Control Model assembles the components of the component model, according to the assembly instructions in the Application Model into the structures de ned in the Implementation Model to create GINA information objects.
The Application Model describes actual GINA applications. Both the description of applications and the GINA and Application Models themselves. These applications are described in terms of components in the Component Model. It represents the set of components that are assembled in order to create GINA information objects as speci ed in the application model.
Ultimately, we have to define GINA applications using a development model that is appropriate for developing GINA applications. And again, the GINA Development Model is itself described as a GINA application.
GINA, at a high level, is a model for modeling. VRDM is a core concept that is embodied by GINA. GINA could be looked at as an environment that turns collected data into a multi-dimensional object environment with each object being connected to other objects through vectors. A key concept of VRDM is that relationships among information objects should themselves be defined as information objects, and be fully configurable.
GINA is a true CBOM for Object Modeling, where the models are themselves executable.
Critical to this approach is the concept of Reflexivity, i.e., the GINA model is described as a GINA model, that permits deep configurability. When one is using the interactive development environment used to create GINA applications, one is using a GINA application. More importantly, GINA assembles GINA applications according to a GINA model for GINA applications. Deeper still, the GINA model is itself an example of a GINA model. Vector Relational Data Modeling Core Concepts
Component Objects
With VRDM, Data Agnostic Objects can be created to represent common relationships called Mechanisms. GINA is designed in that way. If we look back at the concepts associated with GINA, we could say that an object exists in a 3 dimensional data object space. Directory Sub System (DSS)
GINA is implemented through a software-based, multi-layer, configurable data object management environment. Just as the entirety of GINA can be viewed as a series of well-structured layers, the data object management environment is also structured and layered, with multiple layers of the object management environment corresponding to each of the top three layers in the overall GINA. Data Access Layer (DAL)
Task Oriented User Interface (TOUI)
Network architecture means design of computers, devices and media in a network. The computer network architecture can be designed using different ways. File-Server Network:
In File-Server computer network, a powerful computer having disk with large storage capacity and processing power is installed as central computer. This central computer is known as File-Server, Network Server, Application Server or simply a Server. A File-Server stores and manages files. The data files and software are stored on the Server. The individual computers on a network, called nodes, access the data files and software on the Server. Client/Server Network:
Client/Server model is the most popular network model. In client/server network (or arrangement), a powerful computer is used as server. The server controls the functions of network. The software and databases are stored on the server. Different nodes or computers connected to the network can access these software and databases.
All computers (other than computer server) connected in the network are called clients. The clients send requests to the server. Client/Server network may be LAN or WAN.
For example, in a Client/Server network architecture, a database is stored on the Server and the Client computers access the database. The Server portion of the DBMS (Database Management system) is stored on the Server that allows the Clients to add information in database or extract information from database. The Server processes the data and sends the result-to the Client computer.
Peer-To-Peer Network architecture:
In peer-to-peer (P2P) arrangement, all nodes (or computers) on the network have equal status. Each computer stores files on its own storage devices and has its own peripheral devices. A Peer-to-Peer network can also include a Server. In this case, a Peer-to-Peer local area network is similar to a File-Server network architecture.
Hybrid Network Model:
The hybrid network has combined features of both client/server and peer-to-peer network architecture (or arrangements) models. It also has a server.
Network Architecture - Functional architecture for network security systems.The functional network architecture approach to network architecture system design described above is quite general. For s, the security goals are typically derived from the threat analysis. Since the is intended to counter threats uncovered during the threat analysis, each threat should generate a goal involved in countering the threat. For example, if the interfaces with other network subsystems and protocols, goals constraining the design to accommodate the interaction between the and other components are necessary.
For example, threats to may arise from a variety of sources?? passive eavesdroppers, active attackers, etc. Listing each of these as a separate threat might lead to separate functions for to counter each threat. Unless the nature of the threat to is fundamentally different, all threats to should be grouped under the same heading. Fundamental differences between threats within the same class of threat are usually evident when there are basic differences in the security prerequisites, for example, if the pre-provisioned cryptomaterial (keys, passwords, etc.) must be different or if different algorithms must be used.
Sometimes, these differences are generated by backward compatibility requirements necessary to accommodate pre-existing components. After the threats have been, the following steps result in a security of network architecture.
The functional architecture approach to network architecture system design described above is quite general. For s, the security goals are typically derived from the threat analysis. Since the is intended to counter threats uncovered during the threat analysis, each threat should generate a goal involved in countering the threat. Listing each of these as a separate threat might lead to separate functions for to counter each threat. After the threats have been, the following steps result in a security of network architecture.
You may also like:
Computer Networks - What is Computer Networks?
Wide Area Network - What is Wide Area Network?
source: http://howtonetworks.net/network-architecture-what-is-network-architecture
YAMI4 and distributed bioinformatics programs
The situation has come up and will come up again where we would like to spread a task among multiple computers “in parallel”. It’s a cheap and easy way to take a problem that would benefit from not having to be performed sequentially (i.e. comparing one sequence to many others).
A sleek and easy-to-use library for Python, C++, Java and Ada (do people still use Ada?) is YAMI4, at http://inspirel.com/yami4. It allows direct machine-to-machine communication, ideal for creating a control server and spawning some number of host VMs on, say, Amazon’s EC2 cloud, conducting the task in massive parallel, and returning the results.
Code to follow when it’s releasable. You can follow progress on this and other projects at https://github.com/eclarke/saierlab-MHS if you’re really bored.
The Eight Fallacies of Distributed Computing
This is an article I saw on James Gosling’s weblog. As well as pointing you to the source here is the total content as well.
Peter DeutschEssentially everyone, when they first build a distributed application, makes the following eight assumptions. All prove to be false in the long run and all cause big trouble and painful learning experiences.1. The network is reliable
2. Latency is zero
3. Bandwidth is infinite
4. The network is secure
5. Topology doesn’t change
6. There is one administrator
7. Transport cost is zero
8. The network is homogeneous
For more details, read the article by Arnon Rotem-Gal-Oz
What is Adaptinet?
The Adaptinet software platform is the first of its kind. With this platform, we have created a decentralized environment for distributed architectures. Translation: each participant within the network, running a thin, intelligent piece of software can reliably interact with all other members of the network without a central server.
Because Adaptinet relies on intelligent agents at each node rather than servers, there is no upper limit on participation. Thus amazing scale is easily and cost-effectively achievable.
The Distributed Computing SDK
The Adaptinet Distributed Computing SDK is a platform for building distributed applications. At the heart of the SDK is the TransCeiver, which is both a receiver and transmitter of information. The TransCeiver handles all of the basic communications tasks including listening, parsing, message routing and message transmission.
The TransCeiver is a fully threaded application, allowing a number of simultaneous connections with remote nodes. The Adaptinet Distributed Computing SDK infrastructure provides a platform that will easily extend existing applications and give rise to a new class of distributed applications. The SDK is a robust base platform that provides the largest part of the infrastructure needed for developing distributed applications.
