Search Results

Search found 7851 results on 315 pages for 'mobile computing'.

Page 63/315 | < Previous Page | 59 60 61 62 63 64 65 66 67 68 69 70  | Next Page >

  • How/where to run the algorithm on large dataset?

    - by niko
    I would like to run the PageRank algorithm on graph with 4 000 000 nodes and around 45 000 000 edges. Currently I use neo4j graph databse and classic relational database (postgres) and for software projects I mostly use C# and Java. Does anyone know what would be the best way to perform a PageRank computation on such graph? Is there any way to modify the PageRank algorithm in order to run it at home computer or server (48GB RAM) or is there any useful cloud service to push the data along the algorithm and retrieve the results? At this stage the project is at the research stage so in case of using cloud service if possible, would like to use such provider that doesn't require much administration and service setup, but instead focus just on running the algorith once and get the results without much overhead administration work.

    Read the article

  • Cluster Nodes as RAID Drives

    - by BuckWoody
    I'm unable to sleep tonight so I thought I would push this post out VERY early. When you don't sleep your mind takes interesting turns, which can be a good thing. I was watching a briefing today by a couple of friends as they were talking about various ways to arrange a Windows Server Cluster for SQL Server. I often see an "active" node of a cluster with a "passive" node backing it up. That means one node is working and accepting transactions, and the other is not doing any work but simply "standing by" waiting for the first to fail over. The configuration in the demonstration I saw was a bit different. In this example, there were three nodes that were actively working, and a fourth standing by for all three. I've put configurations like this one into place before, but as I was looking at their architecture diagram, it looked familar - it looked like a RAID drive setup! And that's not a bad way to think about your cluster arrangements. The same concerns you might think about for a particular RAID configuration provides a good way to think about protecting your systems in general. So even if you're not staying awake all night thinking about SQL Server clusters, take this post as an opportunity for "lateral thinking" - a way of combining in your mind the concepts from one piece of knowledge to another. You might find a new way of making your technical environment a little better. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Adventures in the Land of CloudDB/NoSQL/NoAcid

    - by KKline
    Cloud, Bunny, or CloudBunny? Last year, some of my friends from Quest Software attended Hadoop World in New York. In 2009, I never would've guessed that Quest would be there with products, community initiatives, as a major sponsor and with presenters? There were just under 1,000 attendees who weren’t the typical devheads and geekasaurs you'd normally see at very techie events like Code Camps, SQL Saturdays, Cloud Camps and or even other NoSQL events such as the Cassandra Summit. We're talkin' enterprise...(read more)

    Read the article

  • Windows Azure Emulators On Your Desktop

    - by BuckWoody
    Many people feel they have to set up a full Azure subscription online to try out and develop on Windows Azure. But you don’t have to do that right away. In fact, you can download the Windows Azure Compute Emulator – a “cloud development environment” – right on your desktop. No, it’s not for production use, and no, you won’t have other people using your system as a cloud provider, and yes, there are some differences with Production Windows Azure, but you’ll be able code, run, test, diagnose, watch, change and configure code without having any connection to the Internet at all. The best thing about this approach is that when you are ready to deploy the code you’ve been testing, a few clicks deploys it to your subscription when you make one.   So what deep-magic does it take to run such a thing right on your laptop or even a Virtual PC? Well, it’s actually not all that difficult. You simply download and install the Windows Azure SDK (you can even get a free version of Visual Studio for it to run on – you’re welcome) from here: http://msdn.microsoft.com/en-us/windowsazure/cc974146.aspx   This SDK will also install the Windows Azure Compute Emulator and the Windows Azure Storage Emulator – and then you’re all set. Right-click the icon for Visual Studio and select “Run as Administrator”:    Now open a new “Cloud” type of project:   Add your Web and Worker Roles that you want to code:   And when you’re done with your design, press F5 to start the desktop version of Azure:   Want to learn more about what’s happening underneath? Right-click the tray icon with the Azure logo, and select the two emulators to see what they are doing:          In the configuration files, you’ll see a “Use Development Storage” setting. You can call the BLOB, Table or Queue storage and it will all run on your desktop. When you’re ready to deploy everything to Windows Azure, you simply change the configuration settings and add the storage keys and so on that you need.   Want to learn more about all this?   Overview of the Windows Azure Compute Emulator: http://msdn.microsoft.com/en-us/library/gg432968.aspx Overview of the Windows Azure Storage Emulator: http://msdn.microsoft.com/en-us/library/gg432983.aspx January 2011 Training Kit: http://www.microsoft.com/downloads/en/details.aspx?FamilyID=413E88F8-5966-4A83-B309-53B7B77EDF78&displaylang=en      

    Read the article

  • Does Azure only support ASP.NET MVC applications and if so how should I adapt my design?

    - by RPK
    I am writing a small ASP.NET Web Application. My worries are that I want to keep the architecture same giving me the option to install it on an Intranet or on a Cloud Platform. I am not using MVC but lately learned that Azure only supports ASP.NET MVC applications. I want to know whether ASP.NET Web Forms application work on Azure/AppHarbor or not. Do I need to convert this application to MVC if Web Forms is not supported? Will the same application run on Intranet as well?

    Read the article

  • Concerns on first ASP.NET cloud application

    - by RPK
    I am writing a small ASP.NET Web Application. My worries are that I want to keep the architecture same giving me the option to install it on an Intranet or on a Cloud Platform. I am not using MVC but lately learned that Azure only supports ASP.NET MVC applications. I want to know whether ASP.NET Web Forms application work on Azure/AppHarbor or not. Do I need to convert this application to MVC if Web Forms is not supported? Will the same application run on Intranet as well?

    Read the article

  • Building massively scalable systems, where to start? [closed]

    - by Mahmoud Hossam
    Recently, I've been seeing these job postings about building scalable systems using Java, and some of the technologies mentioned were: Cassandra Thrift Hadoop MapReduce Among others. How can I get started with these technologies? Is there something else I need to know before actually learning any of these technologies? Maybe some general concepts about building highly available and scalable systems? I already know Java SE, so I won't be starting from scratch.

    Read the article

  • Auto update for application hosted on multiple servers on cloud

    - by mots_g
    I'm working on an application which will run on multiple Amazon EC2 instances. I wish to incorporate auto update feature for my application. The updater should update all the Ec2 instances. Also, there is a central server which governs the creation/termination of EC2 instances as per load. The central server creates a EC2 new instance from a pre-configured custom AMI (custom image which has our application pre-installed). Also, once there is an update, the pre-configured AMI needs to be updated too else it would create new instances which are not updated. Should the central server notify all the ec2 instances for an update and then the instances update themselves?Or should the application on Ec2 instance have a check for periodically updating themselves? Also how should the Amazon custom AMI be updated? Should a new instance be created from it, updated and then a new AMI be re-created and then new images be created from this AMI? What is the best way to incorporate an auto update feature for this architecture? The central server is written in Java and the application running on the cloud is written in C++. Is there a good framework available that can be used for this architecture? Please let me know on what I could be missing in the design and how it would help me to have a nice, extensible and fail safe auto update architecture. Thanks

    Read the article

  • Design pattern for client/server sessions?

    - by nonot1
    Are there any common patterns or general guidance I can learn from for how to design a client/server system where the both the client and server must maintain some kind per-client session state? I've found any number of libraries that can help with some of the plumbing, but it's the overall design I'm wondering about. Open issues in my mind: How to structure the client/server communication so that bidirectional synchronous and asynchronous requests are possible? The server side needs to spawn a couple of per-connected-client session-long helper process. How to manage that? How to manage the mapping from a given client (and any of it's requests) to server state and helper process instances in the face of multiple clients and intermittent network connectivity. Most communication can be simple blocking request/reply, but some will be long running processing tasks that the client will want to keep tabs on. To the extent that it matters, the platform is Linux/C/C++. Not web based. Just an existing thick-client software app being modified to talk to backend servers for some tasks.

    Read the article

  • Launch 7:Windows Phone 7 Style Live Tiles On Android Mobiles

    - by Gopinath
    Android is a great mobile OS but one thought that lingers in the mind of few Android owners is: Am I using a cheap iPhone? This is valid thought for many low end Android users as their phones runs sluggish and the user interface of Android looks like an imitation of iOS. When it comes to Windows Phone 7 users, even though their operating system features are not as great as iPhone/Android but it has its unique user interface; Windows Phone 7 user interface is a very intuitive and fresh, it’s constantly updating Live Tiles show all the required information on the home screen. Android has best mobile operating system features except UI and Windows Phone 7 has excellent user interface. How about porting Windows Phone 7 Tiles interface on an Android? That should be great. Launch 7 app brings the best of Windows Phone 7 look and feel to Android OS. Once the Launcher 7 app is installed and activated, it brings Live Tiles or constantly updating controls that show information on Android home screen. Apart from simple and smooth tiles, there are handful of customization options provided. Users can change colour of the tiles, add new tiles, enable/disable transitions. The reviews on Android Market are on the positive side with 4.4 stars by 10,000 + reviewers. Here are few user reviews 1. Does what it says. only issue for me is that the app drawer doesn’t rotate. And I would like the UI to rotate when my KB is opened. HTC desire z – Jonathan 2. Works great on atrix.Kudos to developers. Awesome. Though needs: Better notification bar More stock images of tiles Better fitting of widgets on tiles – Manny 3. Looks really good like it much more than I thought I would runs real smooth running royal ginger 2.1 – Jay 4. Omg amazing i am definetly keeping it as my default best of android and windows – Devon 5. Man! An update every week! Very very responsive developer! – Andrew You can read more reviews on Android Market here.  There is no doubt that this application is receiving rave reviews. After scanning a while through the reviews, few complaints throw light on the negative side: Battery drains a bit faster & Low end mobile run a bit sluggish. The application is available in two versions – an ad supported free version and $1.41 ad free version. Download Launcher 7 from Android Market This article titled,Launch 7:Windows Phone 7 Style Live Tiles On Android Mobiles, was originally published at Tech Dreams. Grab our rss feed or fan us on Facebook to get updates from us.

    Read the article

  • Building a distributed system on Amazon Web Services

    - by Songo
    Would simply using AWS to build an application make this application a distributed system? For example if someone uses RDS for the database server, EC2 for the application itself and S3 for hosting user uploaded media, does that make it a distributed system? If not, then what should it be called and what is this application lacking for it to be distributed? Update Here is my take on the application to clarify my approach to building the system: The application I'm building is a social game for Facebook. I developed the application locally on a LAMP stack using Symfony2. For production I used an a single EC2 Micro instance for hosting the app itself, RDS for hosting my database, S3 for the user uploaded files and CloudFront for hosting static content. I know this may sound like a naive approach, so don't be shy to express your ideas.

    Read the article

  • Isis Finally Rolls Out

    - by David Dorf
    Google has rolled their wallet out for several chains; I see the NFC readers in Walgreen's when I'm sent their for milk.  But Isis has been relatively quiet until now.  As of last week they have finally launched in their two test cities: Austin, and Salt Lake City.  Below are the supported carriers and phones as of now, but more phones will be added later. Normal 0 false false false EN-US X-NONE X-NONE MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:11.0pt; font-family:"Calibri","sans-serif"; mso-ascii-font-family:Calibri; mso-ascii-theme-font:minor-latin; mso-fareast-font-family:"Times New Roman"; mso-fareast-theme-font:minor-fareast; mso-hansi-font-family:Calibri; mso-hansi-theme-font:minor-latin; mso-bidi-font-family:"Times New Roman"; mso-bidi-theme-font:minor-bidi;} AT&T supports: HTC One™ X, LG Escape™, Samsung Galaxy Exhilarate™, Samsung Galaxy S® III, Samsung Galaxy Rugby Pro™ T-Mobile supports: Samsung Galaxy S® II, Samsung Galaxy S® III, Samsung Galaxy S® Relay 4G Verizon supports: Droid Incredible 4G LTE. Of course iPhone owners have no wallet since Apple didn't included an NFC chip. To start using Isis, you have to take your NFC-capable phone to your carrier's store to get the SIM replaced with a more sophisticated one that has a secure element configured for Isis.  The "secure element" is the cryptographic logic that secures mobile payments.  Carriers like the secure element in the SIM while non-carriers (like Google) prefer the secure element in the phone's electronics. (I'm not entirely sure if you could support both Isis and Google Wallet on the same phone.  Anybody know?) Then you can download the Isis app from Google Play and load your cards.  Most credit cards are supported, and there's a process to verify the credit cards are valid.  Then you can select from the list of participating retailers to "follow."  Selecting a retailer allows that retailer to give you offers via the app. The app is well done and easy to use.  You can select a default payment type and also switch between them easily.  When the phone is tapped on the reader, there are two exchanges of information.  The payment information is transferred, and then the Isis "SmartTap" information which includes optional loyalty number and digital coupons.  Of course the value of mobile wallets comes from the ease of handling all three data types (i.e. payment, loyalty, offers). There are several advertisements for Isis running now, and my favorite is below.

    Read the article

  • Optimistic work sharing on sparsely distributed systems

    - by Asti
    What would a system like BOINC look like if it were written today? At the time BOINC was written, databases were the primary choice for maintaining a shared state and concurrency among nodes. Since then, many approaches have been developed for tasking with optimistic concurrency (OT, partial synchronization primitives, shared iterators etc.) Is there an optimal paradigm for optimistically distributing units of work on sparsely distributing systems which communicate through message passing? Sorry if this is a bit vague. P.S. The concept of Tuple-spaces is great, but locking is inherent to its definition. Edit: I already have a federation system which works very well. I have a reactive OT system is implemented on top of it. I'm looking to extend it to get clients to do units of work.

    Read the article

  • Exclusive use of a Jini server during long-running call

    - by Matthew Flint
    I'm trying to use Jini, in a "Masters/Workers" arrangement, but the Worker jobs may be long running. In addition, each worker needs to have exclusive access to a singleton resource on that machine. As it stands, if a Worker receives another request while a previous request is running, then the new request is accepted and executed in a second thread. Are there any best-practices to ensure that a Worker accepts no further jobs until the current job is complete? Things I've considered: synchronize the job on the server, with a lock on the singleton resource. This would work, but is far from ideal. A call from a Master would block until the current Worker thread completes, even if other Workers become free in the meantime unregister the Worker from the registry while the job is running, then re-register when it completes. Might work OK, but something doesn't smell right with this idea... Of course, I'm quite happy to be told that there are more appropriate technologies than Jini... but I wouldn't want anything too heavyweight, like rolling out EJB containers all over the place.

    Read the article

  • Synchronizing local and remote cache in distributed caching

    - by ltfishie
    With a distributed cache, a subset of the cache is kept locally while the rest is held remotely. In a get operation, if the entry is not available locally, the remote cache will be used and and the entry is added to local cache. In a put operation, both the local cache and remote cache are updated. Other nodes in the cluster also need to be notified to invalidate their local cache as well. What's a simplest way to achieve this if you implemented it yourself, assuming that nodes are not aware of each other. Edit My current implementation goes like this: Each cache entry contains a time stamp. Put operation will update local cache and remote cache Get operation will try local cache then remote cache A background thread on each node will check remote cache periodically for each entry in local cache. If the timestamp on remote is newer overwrite the local. If entry is not found in remote, delete it from local.

    Read the article

  • How should I host our scalable worker processes?

    - by Pieter Breed
    We are designing a new architecture for an enterprise business. The principles we've followed so far is not to develop what you can (possible buy and) deploy, ie, don't reinvent any wheels. In this way we've decided on CQRS, RabbitMQ, Riak and a bunch of other things. We still need to write /some/ business code though and these will be in the form of worker processes, which will consume commands from a message queue and after any side-effects, produce events onto another message queue. The idea behind this is that via the competing-consumers design we will have a scalable design right out of the box. One option is of writing a management infrastructure that will know how to: deploy code instantiate processes kill processes update configuration etc IE provide fault tolerance and scalability. Also, this is exactly what something like GAE and Heroku does for you, but in a public setting and in our organization, public is bad. My question is, is there an out-of-the-box solution that we can use to host our consumers in? Like a private cloud or private platform-as-a-service. Private Heroku or GAE. Is there some kind of software or software product with which we can do all of these things and thereby get scalability and fault tolerance over our consumers?

    Read the article

  • Why do I need two Instances in Windows Azure?

    - by BuckWoody
    Windows Azure as a Platform as a Service (PaaS) means that there are various components you can use in it to solve a problem: Compute “Roles” - Computers running an OS and optionally IIS - you can have more than one "Instance" of a given Role Storage - Blobs, Tables and Queues for Storage Other Services - Things like the Service Bus, Azure Connection Services, SQL Azure and Caching It’s important to understand that some of these services are Stateless and others maintain State. Stateless means (at least in this case) that a system might disappear from one physical location and appear elsewhere. You can think of this as a cashier at the front of a store. If you’re in line, a cashier might take his break, and another person might replace him. As long as the order proceeds, you as the customer aren’t really affected except for the few seconds it takes to change them out. The cashier function in this example is stateless. The Compute Role Instances in Windows Azure are Stateless. To upgrade hardware, because of a fault or many other reasons, a Compute Role's Instance might stop on one physical server, and another will pick it up. This is done through the controlling fabric that Windows Azure uses to manage the systems. It’s important to note that storage in Azure does maintain State. Your data will not simply disappear - it is maintained - in fact, it’s maintained three times in a single datacenter and all those copies are replicated to another for safety. Going back to our example, storage is similar to the cash register itself. Even though a cashier leaves, the record of your payment is maintained. So if a Compute Role Instance can disappear and re-appear, the things running on that first Instance would stop working. If you wrote your code in a Stateless way, then another Role Instance simply re-starts that transaction and keeps working, just like the other cashier in the example. But if you only have one Instance of a Role, then when the Role Instance is re-started, or when you need to upgrade your own code, you can face downtime, since there’s only one. That means you should deploy at least two of each Role Instance not only for scale to handle load, but so that the first “cashier” has someone to replace them when they disappear. It’s not just a good idea - to gain the Service Level Agreement (SLA) for our uptime in Azure it’s a requirement. We point this out right in the Management Portal when you deploy the application: (Click to enlarge) When you deploy a Role Instance you can also set the “Upgrade Domain”. Placing Roles on separate Upgrade Domains means that you have a continuous service whenever you upgrade (more on upgrades in another post) - the process looks like this for two Roles. This example covers the scenario for upgrade, so you have four roles total - One Web and one Worker running the "older" code, and one of each running the new code. In all those Roles you want at least two instances, and this example shows that you're covered for High Availability and upgrade paths: The take-away is this - always plan for forward-facing Roles to have at least two copies. For Worker Roles that do background processing, there are ways to architect around this number, but it does affect the SLA if you have only one.

    Read the article

  • Interconnect nodes in a Java distributed infrastructure for tweet processing

    - by David Moreno García
    I'm working in a new version of an old project that I used to download and process user statuses from Twitter. The main problem of that project was its infrastructure. I used multiple instances of a java application (trackers) to download from Twitter given an specific task (basically terms to search for), connected with a central node (a web application) that had to process all tweets once per day and generate a new task for each trackers once each 15 minutes. The central node also had to monitor all trackers and enable/disable them under user petition. This, as I said, was too slow because I had multiple bottlenecks, so in this new version I want to improve the infrastructure and isolate all functionalities in specific nodes. I also need a good notification system to receive notifications for any node. So, in the next diagram I show the components that I'll need in this new version: As you can see, there are more nodes. Here are some notes about them: Dashboard: Controls trackers statuses and send a single task to each of them (under user request). The trackers will use this task until replaced with a new one (if done, not each 15 minutes like before). Search engine: I need to store all the tweets. They are firstly stored in a local database for each tracker but after that I'm thinking on using something like Elasticsearch to be able to do fast searches. Tweet processor: Just and isolated component with its own database (maybe something like the search engine to have fast access to info generated by the module). In the future more could be added. Application UI: A web application with a shared database with the Dashboard (mainly to store users information and preferences). Indeed, both could be merged into a single web. The main difference with the previous version of the project is that now they will be isolated and they will only show information and send requests. I will not do any heavy task in them (like process tweets as I did before). So, having this components, my main headache is how to structure all to not have to rewrite a lot of code every time I need to access any new data. Another headache is how can I interconnect nodes. I could use sockets but that is a pain in the ass. Maybe a REST layer? And finally, if all the nodes are isolated, how could I generate notifications for each user which info is only in the database used by the Application UI? I'm programming this using Java and Spring (at least I used them in the last version) but I have no problems with changing the language if I can take advantage of a tool/library/engine to make my life easier and have a better platform. Any comment will be appreciated.

    Read the article

  • How to mange big amount users at server side?

    - by Rami
    I built a social android application in which users can see other users around them by gps location. at the beginning thing went well as i had low number of users, But now that I have increasing number of users (about 1500 +100 every day) I revealed a major problem in my design. In my Google App Engine servlet I have static HashMap that holding all the users profiles objects, currenty 1500 and this number will increase as more users register. Why I'm doing it Every user that requesting for the users around him compares his gps with other users and check if they are in his 10km radius, this happens every 5 min on average. That is why I can't get the users from db every time because GAE read/write operation quota will tare me apart. The problem with this desgin is As the number of users increased the Hashmap turns to null every 4-6 hours, I thing that this time is getting shorten but I'm not sure. I'm fixing this by reloading the users from the db every time I detect that it became null, But this causes DOS to my users for 30 sec, So I'm looking for better solution. I'm guessing that it happens because the size of the hashmap, Am I right? I have been advised to use spatial database, but that mean that I can't work with GAE any more and that mean that I need to build my big server all over again and lose my existing DB. Is there something I can do with the existing tools? Thanks.

    Read the article

  • Scalable distributed file system for blobs like images and other documents

    - by Pinnacle
    Cassandra & HBase both do not efficiently support storage of blobs like images. Storing directly on HDFS stresses the Namenode because of huge number of files. Facebook uses Haystack for images and attachments storage, but this is not open source. So is Lustre a good choice for distributed blob storage? I have read that Amazon S3 is used by many, but this would cost money and personally, I would not like to rely on third party system. What are other suggestions?

    Read the article

  • Is there a canonical resource on multi-tenancy web applications using ruby + rails

    - by AlexC
    Is there a canonical resource on multi-tenancy web applications using ruby + rails. There are a number of ways to develop rails apps using cloud capabilities with real elastic properties but there seems to be a lack of clarity with how to achieve multitenancy, specifically at the model / data level. Is there a canonical resource on options to developing multitenancy rails applications with the required characteristics of data seperation, security, concurrency and contention required by an enterprise level cloud application.

    Read the article

  • Erlang node acts like it connects, but doesn't [migrated]

    - by Malfist
    I'm trying to setup a distributed network of nodes across a few firewalls and it's not going so well. My application is structured like this: there is a central server that always running a node ([email protected]) and my co-worker's laptops connect to it on startup. This works if we're all in the office, but if someone is at home, they can connect to the masternode, but they fail to connect to the other nodes in the swarm. I.E., erlang fails to gossip correctly. To correct this, I've change epmd's port number and changed the inet_dist_listen ports to a known open port (1755 and 7070 respectively). However, something fishy is going on. I can run net_adm:world() and it reports that it connects to master node, but when I run nodes() I get an empty array. Same with net_adm:ping('[email protected]'). See: Eshell V5.9 (abort with ^G) ([email protected])1> net_adm:world(). ['[email protected]'] ([email protected])2> nodes(). [] ([email protected])3> net_adm:ping('[email protected]'). pong ([email protected])4> nodes(). [] ([email protected])5> What's going on, and how can I fix it?

    Read the article

  • High Availability for IaaS, PaaS and SaaS in the Cloud

    - by BuckWoody
    Outages, natural disasters and unforeseen events have proved that even in a distributed architecture, you need to plan for High Availability (HA). In this entry I'll explain a few considerations for HA within Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS). In a separate post I'll talk more about Disaster Recovery (DR), since each paradigm has a different way to handle that. Planning for HA in IaaS IaaS involves Virtual Machines - so in effect, an HA strategy here takes on many of the same characteristics as it would on-premises. The primary difference is that the vendor controls the hardware, so you need to verify what they do for things like local redundancy and so on from the hardware perspective. As far as what you can control and plan for, the primary factors fall into three areas: multiple instances, geographical dispersion and task-switching. In almost every cloud vendor I've studied, to ensure your application will be protected by any level of HA, you need to have at least two of the Instances (VM's) running. This makes sense, but you might assume that the vendor just takes care of that for you - they don't. If a single VM goes down (for whatever reason) then the access to it is lost. Depending on multiple factors, you might be able to recover the data, but you should assume that you can't. You should keep a sync to another location (perhaps the vendor's storage system in another geographic datacenter or to a local location) to ensure you can continue to serve your clients. You'll also need to host the same VM's in another geographical location. Everything from a vendor outage to a network path problem could prevent your users from reaching the system, so you need to have multiple locations to handle this. This means that you'll have to figure out how to manage state between the geo's. If the system goes down in the middle of a transaction, you need to figure out what part of the process the system was in, and then re-create or transfer that state to the second set of systems. If you didn't write the software yourself, this is non-trivial. You'll also need a manual or automatic process to detect the failure and re-route the traffic to your secondary location. You could flip a DNS entry (if your application can tolerate that) or invoke another process to alias the first system to the second, such as load-balancing and so on. There are many options, but all of them involve coding the state into the application layer. If you've simply moved a state-ful application to VM's, you may not be able to easily implement an HA solution. Planning for HA in PaaS Implementing HA in PaaS is a bit simpler, since it's built on the concept of stateless applications deployment. Once again, you need at least two copies of each element in the solution (web roles, worker roles, etc.) to remain available in a single datacenter. Also, you need to deploy the application again in a separate geo, but the advantage here is that you could work out a "shared storage" model such that state is auto-balanced across the world. In fact, you don't have to maintain a "DR" site, the alternate location can be live and serving clients, and only take on extra load if the other site is not available. In Windows Azure, you can use the Traffic Manager service top route the requests as a type of auto balancer. Even with these benefits, I recommend a second backup of storage in another geographic location. Storage is inexpensive; and that second copy can be used for not only HA but DR. Planning for HA in SaaS In Software-as-a-Service (such as Office 365, or Hadoop in Windows Azure) You have far less control over the HA solution, although you still maintain the responsibility to ensure you have it. Since each SaaS is different, check with the vendor on the solution for HA - and make sure you understand what they do and what you are responsible for. They may have no HA for that solution, or pin it to a particular geo, or perhaps they have a massive HA built in with automatic load balancing (which is often the case).   All of these options (with the exception of SaaS) involve higher costs for the design. Do not sacrifice reliability for cost - that will always cost you more in the end. Build in the redundancy and HA at the very outset of the project - if you try to tack it on later in the process the business will push back and potentially not implement HA. References: http://www.bing.com/search?q=windows+azure+High+Availability  (each type of implementation is different, so I'm routing you to a search on the topic - look for the "Patterns and Practices" results for the area in Azure you're interested in)

    Read the article

< Previous Page | 59 60 61 62 63 64 65 66 67 68 69 70  | Next Page >