Search Results

Search found 5597 results on 224 pages for 'restful architecture'.

Page 46/224 | < Previous Page | 42 43 44 45 46 47 48 49 50 51 52 53  | Next Page >

• xmpp flow -server, client and library

  - by Him

  My complete requirement is development of a chat engine - including server, clients etc. Currently I am working on things at my desktop only but once done, I have to host it; basically incorporate it with in a site for chatting purpose. So, now my problem is: I am not clear about how the actual data flow is? I have googled and read about xmpp (a book by Peter Andre) also but I am not clear about the flow and what are the actual requirements to do the above mentioned task. What I currently know is: 1) I need a server - so selected ejabberd 2) I need client - still not sure which one to use and one other doubt is how this client thing will work when deployed on some website for chatting purpose. 3) Some library - don't know which one and what is the purpose? Can anyone guide me?

  Read the article

• Sharing VBO with multiple objects and fixed size buffer data

  - by Mark Ingram

  I'm just messing around with OpenGL and getting some basic structures in place and my first attempt resulted in each SceneObject class (just contains vertex information right now) having it's own VBO inside it, however I've read that it might be better to share VBOs across multiple objects. Also, I read that you should avoid resizing a VBO (repeated calls to glBufferData with different size parameters), and instead choose a fixed size for a VBO, and just try a range from the buffer. I don't think changing the size of the buffer data would happen too often, but surely it would be better to only allocate the data you need? Choosing an arbitrary value seems risky. I'm looking for some advice on working with individual objects in a scene and their associated buffer data.

  Read the article

• Clustering and custom applications

  - by Ahmed ilyas

  I was not entirely sure what tags to put but hope this is ok. This is just a general question in regards to clustering and applications: so lets say we have a clustered environment setup. We cluster SQL Server (I dont know exactly how its done but lets just say its been done for the sake of argument). Now if a website or application is trying to access that database for read/write (say an ASP.NET app or a C# Winforms app) and during that time SQL goes down - it takes a couple of minutes for the clustering failover to take affect to switch to another node. What happens during this time? I think it will time out/unable to connect. BUT is there a way for it to place the request in some pipeline so when the cluster node is back up/switched over it will continue as normal? as you can see, I know nothing much about clustering! what about your own custom .NET apps? Would there be a special way to develop them? I know that you can say create a simple Hello world app, and cluster that but they wouldnt be something you could see interms of the UI or anything, so they would effectively need to be developed as a Windows Service perhaps or even as a standard Console app which runs and not wait for user input but you wouldnt see any output from it (unless you redirect output to somewhere else) What im getting at here is... for those who have experience or developed a cluster application in .NET, how did you do it and what are the things to be aware of? For example we have the cloud service - fundamentally its built on clustering - if there is an outage, another node takes place and service is resumed as normal but we dont really see much of that downtime.

  Read the article

• Low-level game engine renderer design

  - by Mark Ingram

  I'm piecing together the beginnings of an extremely basic engine which will let me draw arbitrary objects (SceneObject). I've got to the point where I'm creating a few sensible sounding classes, but as this is my first outing into game engines, I've got the feeling I'm overlooking things. I'm familiar with compartmentalising larger portions of the code so that individual sub-systems don't overly interact with each other, but I'm thinking more of the low-level stuff, starting from vertices working up. So if I have a Vertex class, I can combine that with a list of indices to make a Mesh class. How does the engine determine identical meshes for objects? Or is that left to the level designer? Once we have a Mesh, that can be contained in the SceneObject class. And a list of SceneObject can be placed into the Scene to be drawn. Right now I'm only using OpenGL, but I'm aware that I don't want to be tying OpenGL calls right in to base classes (such as updating the vertices in the Mesh, I don't want to be calling glBufferData etc). Are there any good resources that discuss these issues? Are there any "common" heirachies which should be used?

  Read the article

• Running multiple box2D world objects on a server

  - by CharbelAbdo

  I'm creating a multiplayer game using LibGdx (with Box2d) and Kryonet. Since this is the first time I work on multiplayer games, I read a bit about server - client implementations, and it turns out that the server should handle important tasks like collision detection, hits, characters dying etc... Based on some articles (like the excellent Gabriel Gambetta Fast paced multiplayer series), I also know that the client should work in parallel to avoid the lag while the server responds to commands. Physics wise, each game will have 2 players, and any projectiles fired. What I'm thinking of doing is the following: Create a physics world on the client When the game is signaled to start, I create the same physics world on the server (without any rendering obviously). Whenever the player issues a command (move or fire), I send the command to the server and immediately start processing it on the client. When the server receives the command, it applies it on the server's world (set velocity etc...) Each 100ms, the server sends the new state to the client which corrects what was calculated locally. Any critical action (hit, death, level up) is calculated only on the server and sent to the client. Essentially, I would have a Box2d World object running on the server for each game in progress, in sync with the worlds running on the clients. The alternative would be to do my own calculations on the server instead of relying on Box2D to do them for me, but I'm trying to avoid that. My question is: Is it wise to have, for example, 1000 instances of the World object running and executing steps on the server? Tomcat used around 750 MBytes of memory when trying it without any object added to the world. Anybody tried that before? If not, is there any alternative? Google did not help me, are there any guidelines to use when you want to have physics on both the client and the server? Thanks for any help.

  Read the article

• Design Patterns and their most common uses for them [closed]

  - by cable729

  Possible Duplicate: What are some programming design patterns that are useful in game development? As I'm returning to game dev, I've realized that I've lost a lot of the knowledge I had before. So now I'm looking at design patterns that I can use for my next project. One design pattern that I've seen a lot is the 'composition method,' which uses actors and components. Is that the right name for it? I'd like to look more at this and see what the advantages/pitfalls are. So what design patterns are out there, and what are the advantages/disadvantages to them?

  Read the article

• How to create an Orthographic display in OpenGL (ES) that handles different screen sizes and orientations?

  - by Piku

  I'm trying to create an iPad/iPhone game using GLES2.0 that contains a 3D scene with a heads-up-display/GUI overlaid on the top. However, this problem would also apply if I were to port my game to a computer and run the game in a resizable window, or allow the user to change screen resolutions... When trying to make the 2D GUI/HUD work I've made the assumption that all I'm really doing is drawing a load of 2D textured 'quads' on the screen and am trying to treat the orthographic projection as an old-style 2D display with 0,0 in the upper left and screenWidth,ScreenHeight in the lower right. This causes me all sorts of confusion when I rotate my ipad into Landscape mode since I can't work out what to put into my projection and modelview matrices to turn everything around the right way. It also gets messy if I want to support the iPad's large screen, an iPhone or a Retina display since I have to then draw three sets of textures for everything and work out which ones to use. Should I be trying to map the 2D OpenGL co-ords 1:1 with the screen? While typing out this question it occurs to me that I could keep my origin in the centre, still running -1/+1 along the axes. This would let me scale my 2D content appropriately on the different screen sizes, but wouldn't I end up with the textures being scaled and possibly losing quality? I'm using OpenGLES 2.0 and have a matrix library that has equivalents to the GLES1.1 glOrthof() and glFrustrum() calls.

  Read the article

• Role of an entity state in a component based system?

  - by Paul

  Component-based entity systems are all the rage these days; everyone seems to agree they are the way to go, but no one really has a definitive implementation of such a system. I was wondering, what role do entity states (walking-left, standing, jumping, etc) have in a CBS? Do they act like controllers (i.e. they handle events and change the entity's attributes based on those events)? What about cases where a state would, for example, require that the entity enters no-clip mode? Should, that state, when it enters, maybe set the CollisionComponent of the entity to a null pointer or something? (Then, on exit, the state should restore the entity's CollisionComponent to its previous state.) Also, I guess it's the current state's job to change the entity's state to something else, right?

  Read the article

• Ensuring non conflicting components in a modular system

  - by Hailwood

  So lets say we are creating a simple "modular system" framework. The bare bones might be the user management. But we want things like the Page Manager, the Blog, the Image Gallery to all be "optional" components. So a developer could install the Page Manager to allow their client to add a static home page and about page with content they can easily edit with a wysiwyg editor. The developer could then also install the Blog component to allow the client to add blog entries. The developer could then also install the Gallery component to allow the client to show off a bunch of images. The thing is, all these components are designed to be independent, so how do we go about ensuring they don't clash? E.g. ensuring the client doesn't create a /gallery page with the Page Manager and then wonder why the gallery stopped working, or the same issue with the Blog component, assuming we allow the users to customize the URL structure of the blog (because remember, the Page Manager doesn't necessarily have to be there, so we might not wan't our blog posts to be Date/Title formatted), likewise our clients aren't always going to be happy to have their pages under pages/title formatting. My core question here is, when building a modular system how to we ensure that the modules don't conflict without restricting functionality? Do we just leave it up to the clients/developer using the modules to ensure they get setup in a way that does not conflict?

  Read the article

• Is having functionality in DB a road block to scalability?

  - by Estefany Velez

  I may not be able to give the right title to the question. But here it is, We are developing financial portal for wealth management. We are expecting over 10000 clients to use the application. The portal calculates various performance analytics based on the the technical analysis of the stock market. We developed lot of the functionality through Stored procedures, user defined functions, triggers etc. through Database. We thought we can gain huge performance boost doing stuff directly in database than through C# code. And we actually did get a huge performance boost. When I tried to brag about the achievement to our CTO, he counter questioned my decision of having functionality implemented in database rather than code. According to him such applications suffer scalability problems. In his words "These days things are kept in memory/cache. Clustered data is hard to manage over time. Facebook, Google have nothing in database. It is the era of thin servers and thick clients. DB is used only to store plain data and functionality should be completely decoupled from the database." Can you guys please give me some suggestions as to whether what he says is right. How to go about architect such an application?

  Read the article

• Is component-based design an architectural pattern or design pattern?

  - by xEnOn

  When using the component-based paradigm in game development with engines like Unity, is component-based design an architectural pattern, or a design pattern? Can I even say that component-based design is my "main" architectural pattern for my game? I see architectural patterns as being more high-level than design pattern. The component-based design in game development's context (like with Unity engine) seems to fit as an architectural pattern to me. However, on some sites, I read that component-based design is a behavioural pattern, much like other behavioural design patterns, and not so much like an architectural pattern like MVC.

  Read the article

• Multiplayer game communication framework for mac/ios

  - by ishaq

  (Cross post from stackoverflow) I am creating a multiplayer 2D game for Mac and iOS devices. I'll be using cocso2d for graphics/game engine, however I am largely blank on what to use for multiplayer communication. Please note that I cannot use central severs e.g. SmartFox, RedDwarf, etc since I want the players to "host" games for others and be able to play it on their LAN, VPN or my own servers. Any pointers? I checked lidgren but it's for .NET only and hence not an option for me. EDIT: just in case it wasn't clear, the messaging has to be real time hence it's probably going to be over UDP

  Read the article

• Why should I adopt MVC?

  - by Andrew

  I decided to get my hands wet and got the YII framework for PHP. I created my first application, then created new controller, model and view. Connected to database, got my record passed from controller to the view and printed the hello world. I am confused now. If I have to do the same thing for each page, this seems like a nightmare to me. In each controller I have to do a lot of same operations - declare variables, and pass them to views. I also need to create models for each page and this is all confusing to me. In my idea the main goal of development is to avoid duplication, but what I see here is lots and lots of duplicated code. Please advise and clarify. Maybe you could suggest a good reading about MVC and coding patterns and best practices in MVC. Because so far, it takes much more time to create a small site using MVC than using my own programming schema.

  Read the article

• Event Driven Communication in Game Engine - Yes or No?

  - by Bunkai.Satori

  As I am reading book Game Coding Complete (http://www.amazon.com/Game-Coding-Complete-Third-McShaffry/dp/1584506806/ref=sr_1_1?ie=UTF8&qid=1295978774&sr=8-1), the author recommend Event Driven communication among the all game objects and modules. Basicaly, all the living game actors and object should communicate with the key modules (Physics, AI, Game Logic, Game View, etc..) via internal event messaging system. This would mean designing efficient event manager as well. My question is, whether this is proven and recommended approach. If it is not properly designed, it might mean consuming a lot of CPU cycles, which can be used elsewhere. This is especially true, if the game is targetted for mobile platform. What is your opinion and recommendation, please?

  Read the article

• Managing multiple references of the same game entity in different places using IDs

  - by vargonian

  I've seen great questions on similar topics, but none that addressed this particular method: Given that I have multiple collections of game entities in my [XNA Game Studio] game, with many entities belonging to multiple lists, I'm considering ways I could keep track of whenever an entity is destroyed and remove it from the lists it belongs to. A lot of potential methods seem sloppy/convoluted, but I'm reminded of a way I've seen before in which, instead of having multiple collections of game entities, you have collections of game entity IDs instead. These IDs map to game entities via a central "database" (perhaps just a hash table). So, whenever any bit of code wants to access a game entity's members, it first checks to see if it's even in the database still. If not, it can react accordingly. Is this a sound approach? It seems that it would eliminate many of the risks/hassles of storing multiple lists, with the tradeoff being the cost of the lookup every time you want to access an object.

  Read the article

• How do I create my own programming language and a compiler for it

  - by Dave

  I am thorough with programming and have come across languages including BASIC, FORTRAN, COBOL, LISP, LOGO, Java, C++, C, MATLAB, Mathematica, Python, Ruby, Perl, Javascript, Assembly and so on. I can't understand how people create programming languages and devise compilers for it. I also couldn't understand how people create OS like Windows, Mac, UNIX, DOS and so on. The other thing that is mysterious to me is how people create libraries like OpenGL, OpenCL, OpenCV, Cocoa, MFC and so on. The last thing I am unable to figure out is how scientists devise an assembly language and an assembler for a microprocessor. I would really like to learn all of these stuff and I am 15 years old. I always wanted to be a computer scientist some one like Babbage, Turing, Shannon, or Dennis Ritchie. I have already read Aho's Compiler Design and Tanenbaum's OS concepts book and they all only discuss concepts and code in a high level. They don't go into the details and nuances and how to devise a compiler or operating system. I want a concrete understanding so that I can create one myself and not just an understanding of what a thread, semaphore, process, or parsing is. I asked my brother about all this. He is a SB student in EECS at MIT and hasn't got a clue of how to actually create all these stuff in the real world. All he knows is just an understanding of Compiler Design and OS concepts like the ones that you guys have mentioned (ie like Thread, Synchronisation, Concurrency, memory management, Lexical Analysis, Intermediate code generation and so on)

  Read the article

• Broad topics needed for teaching game development

  - by livingtech

  I am going to be doing a presentation on game development to an iPhone user group in the near(ish) future. My audience are iPhone developers, but not necessarily very experienced ones, and this is meant to be an introduction. My question is, what broad topics are needed to understand game development? I acknowledge that this is fairly subjective, but I really am hoping for a comprehensive list of high-level topics that apply to a broad enough swath of games that anyone interested in the topic SHOULD know about them. I would be ecstatic with some pointers to any resources that attempt to make a list such as this this. (I have looked, but my google-fu is failing me tonight.) Here's what I have so far: The Game Loop a sub-note about event driven games 2D Animation sprites/texture maps 3D Animation importance of frameworks modeling software Particles and particle effects hit detection AI Obviously I will not be covering all these topics with any depth, more like simply defining them so that after my talk, the audience will (hopefully) be able to wrap their heads around how any given game might be developed. What am I missing?

  Read the article

• Rendering order in an Entity System

  - by Daedalus

  Say I use a basic ES approach, and also inside Systems I hold lists of all entities that Systems are required to process. How do I maintain this list of entities in desired rendering order, i.e. for a dumb 2D RenderingSystem? I saw this discussion, and what they suggest is to do something like Z ordering - what I would probably do is just to store a "layer" int in DrawableComponent and then, inside RenderingSystem, just sort entities by mentioned "layer" whenever the entity list for RenderingSystem changes. They also say we could just delete and recreate the entity whenever we want it on the top, but it seems too inflexible to me. How is this problem usually solved?

  Read the article

• Where should you put constants and why?

  - by Tim Meyer

  In our mostly large applications, we usually have a only few locations for constants: One class for GUI and internal contstants (Tab Page titles, Group Box titles, calculation factors, enumerations) One class for database tables and columns (this part is generated code) plus readable names for them (manually assigned) One class for application messages (logging, message boxes etc) The constants are usually separated into different structs in those classes. In our C++ applications, the constants are only defined in the .h file and the values are assigned in the .cpp file. One of the advantages is that all strings etc are in one central place and everybody knows where to find them when something must be changed. This is especially something project managers seem to like as people come and go and this way everybody can change such trivial things without having to dig into the application's structure. Also, you can easily change the title of similar Group Boxes / Tab Pages etc at once. Another aspect is that you can just print that class and give it to a non-programmer who can check if the captions are intuitive, and if messages to the user are too detailed or too confusing etc. However, I see certain disadvantages: Every single class is tightly coupled to the constants classes Adding/Removing/Renaming/Moving a constant requires recompilation of at least 90% of the application (Note: Changing the value doesn't, at least for C++). In one of our C++ projects with 1500 classes, this means around 7 minutes of compilation time (using precompiled headers; without them it's around 50 minutes) plus around 10 minutes of linking against certain static libraries. Building a speed optimized release through the Visual Studio Compiler takes up to 3 hours. I don't know if the huge amount of class relations is the source but it might as well be. You get driven into temporarily hard-coding strings straight into code because you want to test something very quickly and don't want to wait 15 minutes just for that test (and probably every subsequent one). Everybody knows what happens to the "I will fix that later"-thoughts. Reusing a class in another project isn't always that easy (mainly due to other tight couplings, but the constants handling doesn't make it easier.) Where would you store constants like that? Also what arguments would you bring in order to convince your project manager that there are better concepts which also comply with the advantages listed above? Feel free to give a C++-specific or independent answer. PS: I know this question is kind of subjective but I honestly don't know of any better place than this site for this kind of question. Update on this project I have news on the compile time thing: Following Caleb's and gbjbaanb's posts, I split my constants file into several other files when I had time. I also eventually split my project into several libraries which was now possible much easier. Compiling this in release mode showed that the auto-generated file which contains the database definitions (table, column names and more - more than 8000 symbols) and builds up certain hashes caused the huge compile times in release mode. Deactivating MSVC's optimizer for the library which contains the DB constants now allowed us to reduce the total compile time of your Project (several applications) in release mode from up to 8 hours to less than one hour! We have yet to find out why MSVC has such a hard time optimizing these files, but for now this change relieves a lot of pressure as we no longer have to rely on nightly builds only. That fact - and other benefits, such as less tight coupling, better reuseability etc - also showed that spending time splitting up the "constants" wasn't such a bad idea after all ;-)

  Read the article

• Develop for Desktop and mobile use?

  - by ran2

  I am in the very beginning of developing an app / desktop program. I want it to be cross-platform and possibly also as a tablet version (preferably Android Icecream sandwich). Note that I need to run it offline. I thought about the following approaches: ADOBE Air, since I do not need much performance. Plus I did some web programming in the past which might be of some use. Afaik it would run on OS X and Windows and should run on mobile OSes, too. Qt. Found some nice Qt based desktop recently and read it also works on android. Plus I like the SDK. HTML5 / JS. Again my web background should help me here. I wont need no sever side scripts, thus it should work without installing anything but a browser. How easy could this be converted into an Android app? There might be a plethora of other (better) ways to do it, but I haven't thought of them yet. Can you help out? How would you create such an application. Would it be better to do some pure desktop client and then create tablet versions? Would you rather start to create a website and worry later on how to turn into an app?

  Read the article

• Questions about identifying the components in MVC

  - by luiscubal

  I'm currently developing an client-server application in node.js, Express, mustache and MySQL. However, I believe this question should be mostly language and framework agnostic. This is the first time I'm doing a real MVC application and I'm having trouble deciding exactly what means each component. (I've done web applications that could perhaps be called MVC before, but I wouldn't confidently refer to them as such) I have a server.js that ties the whole application together. It does initialization of all other components (including the database connection, and what I think are the "models" and the "views"), receiving HTTP requests and deciding which "views" to use. Does this mean that my server.js file is the controller? Or am I mixing code that doesn't belong there? What components should I break the server.js file into? Some examples of code that's in the server.js file: var connection = mysql.createConnection({ host : 'localhost', user : 'root', password : 'sqlrevenge', database : 'blog' }); //... app.get("/login", function (req, res) { //Function handles a GET request for login forms if (process.env.NODE_ENV == 'DEVELOPMENT') { mu.clearCache(); } session.session_from_request(connection, req, function (err, session) { if (err) { console.log('index.js session error', err); session = null; } login_view.html(res, user_model, post_model, session, mu); //I named my view functions "html" for the case I might want to add other output types (such as a JSON API), or should I opt for completely separate views then? }); }); I have another file that belongs named session.js. It receives a cookies object, reads the stored data to decide if it's a valid user session or not. It also includes a function named login that does change the value of cookies. First, I thought it would be part of the controller, since it kind of dealt with user input and supplied data to the models. Then, I thought that maybe it was a model since it dealt with the application data/database and the data it supplies is used by views. Now, I'm even wondering if it could be considered a View, since it outputs data (cookies are part of HTTP headers, which are output)

  Read the article

• About shared (static) Members and its behavior

  - by Allende

  I just realized that I can access shared members from instances of classes (probably this is not correct, but compile and run), and also learn/discover that, I can modify shared members, then create a new instance and access the new value of the shared member. My question is, what happens to the shared members, when it comes back to the "default" value (class declaration), how dangerous is it do this ? is it totally bad ? is it valid in some cases ?. If you want to test my point here is the code (console project vb.net) that I used to test shared members, as you can see/compile/run, the shared member "x" of the class "Hello" has default value string "Default", but at runtime it changes it, and after creating a new object of that class, this object has the new value of the shared member. Module Module1 Public Class hello Public Shared x As String = "Default" Public Sub New() End Sub End Class Sub Main() Console.WriteLine("hello.x=" & hello.x) Dim obj As New hello() Console.WriteLine("obj.x=" & obj.x) obj.x = "Default shared memeber, modified in object" Console.WriteLine("obj.x=" & obj.x) hello.x = "Defaul shared member, modified in class" Console.WriteLine("hello.x=" & hello.x) Dim obj2 As New hello() Console.WriteLine("obj2.x=" & obj2.x) Console.ReadLine() End Sub End Module UPDATE: First at all, thanks to everyone, each answer give feedback, I suppose, by respect I should choose one as "the answer", I don't want to be offensive to anyone, so please don't take it so bad if I didn't choose you answer.

  Read the article

• From a DDD perspective is a report generating service a domain service or an infrastructure service?

  - by Songo

  Let assume we have the following service whose responsibility is to generate Excel reports: class ExcelReportService{ public String generateReport(String fileFormatFilePath, ResultSet data){ ReportFormat reportFormat = new ReportFormat(fileFormatFilePath); ExcelDataFormatterService excelDataFormatterService = new ExcelDataFormatterService(); FormattedData formattedData = excelDataFormatterService.format(data); ExcelFileService excelFileService = new ExcelFileService(); String reportPath= excelFileService.generateReport(reportFormat,formattedData); return reportPath; } } This is pseudo code for the service I want to design where: fileFormatFilePath: path to a configuration file where I'll keep the format of my excel file (headers, column widths, number of columns,..etc) data: the actual records returned from the database. This data can't be used directly coz I might need to make further calculations to the data before inserting them to the excel file. ReportFormat: Value object to hold the report format, has methods like getHeaders(), getColumnWidth(),...etc. ExcelDataFormatterService: a service to hold any logic that need to be applied to the data returned from the database before inserting it to the file. FormattedData: Value object the represents the formatted data to be inserted. ExcelFileService: a wrapper top the 3rd party library that generates the excel file. Now how do you determine whether a service is an infrastructure or domain service? I have the following 3 services here: ExcelReportService, ExcelDataFormatterService and ExcelFileService?

  Read the article

• Is this project Structure Valid?

  - by rafuru

  I have a dilemma: In the university we learn to create modular software (on java), but this modularity is explained using a single project with packages (a package for business, another one for DAOS and another one for the model, oh and a last package for frontend). But in my work we use the next structure: I will try to explain: First we create a java library project where the model (entities classes) are created in a package. Next we create an EJB named DAOS and using the netbeans wizard we store the DAOS interfaces in the library project in another package , these interfaces are implemented in the DAOS bean. So the next part is the business logic, we create a business EJB for each group of functions , again using the wizard we store the interface in the java library project in another package then is implemented on the business bean. The final part (for the backend) is a bean that I have suggested: a Facade bean who will gather every method of the business beans in a single bean and this has an interface too that is created in our library project and implemented in the bean. So the next part is call the facade module on the web project. But I don't know how valid or viable is this, maybe I'm doing everything wrong and I don't even know! so I want to ask your opinion about this.

  Read the article

• How can I gather client's data on Google App Engine without using Datastore/Backend Instances too much?

  - by ruslan

  One of the projects I'm working on is online survey engine. It's my first big commercial project on Google App Engine. I need your advice on how to collect stats and efficiently record them in DataStore without bankrupting me. Initial requirements are: After user finishes survey client sends list of pairs [ID (int) + PercentHit (double)]. This list shows how close answers of this user match predefined answers of reference answerers (which identified by IDs). I call them "target IDs". Creator of the survey wants to see aggregated % for given IDs for last hour, particular timeframe or from the beginning of the survey. Some surveys may have thousands of target/reference answerers. So I created entity public class HitsStatsDO implements Serializable { @Id transient private Long id; transient private Long version = (long) 0; transient private Long startDate; @Parent transient private Key parent; // fake parent which contains target id @Transient int targetId; private double avgPercent; private long hitCount; } But writing HitsStatsDO for each target from each user would give a lot of data. For instance I had a survey with 3000 targets which was answered by ~4 million people within one week with 300K people taking survey in first day. Even if we assume they were answering it evenly for 24 hours it would give us ~1040 writes/second. Obviously it hits concurrent writes limit of Datastore. I decided I'll collect data for one hour and save that, that's why there are avgPercent and hitCount in HitsStatsDO. GAE instances are stateless so I had to use dynamic backend instance. There I have something like this: // Contains stats for one hour private class Shard { ReadWriteLock lock = new ReentrantReadWriteLock(); Map<Integer, HitsStatsDO> map = new HashMap<Integer, HitsStatsDO>(); // Key is target ID public void saveToDatastore(); public void updateStats(Long startDate, Map<Integer, Double> hits); } and map with shard for current hour and previous hour (which doesn't stay here for long) private HashMap<Long, Shard> shards = new HashMap<Long, Shard>(); // Key is HitsStatsDO.startDate So once per hour I dump Shard for previous hour to Datastore. Plus I have class LifetimeStats which keeps Map<Integer, HitsStatsDO> in memcached where map-key is target ID. Also in my backend shutdown hook method I dump stats for unfinished hour to Datastore. There is only one major issue here - I have only ONE backend instance :) It raises following questions on which I'd like to hear your opinion: Can I do this without using backend instance ? What if one instance is not enough ? How can I split data between multiple dynamic backend instances? It hard because I don't know how many I have because Google creates new one as load increases. I know I can launch exact number of resident backend instances. But how many ? 2, 5, 10 ? What if I have no load at all for a week. Constantly running 10 backend instances is too expensive. What do I do with data from clients while backend instance is dead/restarting?

  Read the article

< Previous Page | 42 43 44 45 46 47 48 49 50 51 52 53  | Next Page >