Search Results

Search found 8688 results on 348 pages for 'per salmi'.

Page 60/348 | < Previous Page | 56 57 58 59 60 61 62 63 64 65 66 67  | Next Page >

  • SOA performance on SPARC T5 benchmark results

    - by JuergenKress
    The brand NEW super fast SPARC T5 servers are available. The platform is superb to run large SOA Suite environments or to consolidate your whole middleware platform. Some performance advices, recommended for all workloads: Performance profile for SOA apps on Oracle Solaris 11 BPEL (Fusion Order Demo) instances per second OSB (messages / transformations per second) Crypto acceleration study for SOA transformations SPARC T4 and T5 platform testing, pre-tuning Performance suitable for mid-to-high range enterprise in stand-alone SOA deployment or virtualized consolidation environment shared with Oracle applications 2.2x to 5x faster than SPARC T3 servers 25% faster SOA throughput, core to core than Intel 5600-series servers (running Exalogic software) SPARC T5 has 2x the consolidation density of Intel 5600-class processors 2x faster initial deployment time using Optimized Solutions pre-tested configuration steps Over 200 Application adapters for easiest Oracle software integration Would you like to get details? We can share with you on 1:1 bases T5 SOA Suite performance benchmarks, please contact your local partner manager or myself! SOA & BPM Partner Community For regular information on Oracle SOA Suite become a member in the SOA & BPM Partner Community for registration please visit www.oracle.com/goto/emea/soa (OPN account required) If you need support with your account please contact the Oracle Partner Business Center. Blog Twitter LinkedIn Facebook Wiki Mix Forum Technorati Tags: T5,TS Sparc,T5 SOA,bechmark,SOA Community,Oracle SOA,Oracle BPM,Community,OPN,Jürgen Kress

    Read the article

  • Processing Kinect v2 Color Streams in Parallel

    - by Chris Gardner
    Originally posted on: http://geekswithblogs.net/freestylecoding/archive/2014/08/20/processing-kinect-v2-color-streams-in-parallel.aspxProcessing Kinect v2 Color Streams in Parallel I've really been enjoying being a part of the Kinect for Windows Developer's Preview. The new hardware has some really impressive capabilities. However, with great power comes great system specs. Unfortunately, my little laptop that could is not 100% up to the task; I've had to get a little creative. The most disappointing thing I've run into is that I can't always cleanly display the color camera stream in managed code. I managed to strip the code down to what I believe is the bear minimum: using( ColorFrame _ColorFrame = e.FrameReference.AcquireFrame() ) { if( null == _ColorFrame ) return;   BitmapToDisplay.Lock(); _ColorFrame.CopyConvertedFrameDataToIntPtr( BitmapToDisplay.BackBuffer, Convert.ToUInt32( BitmapToDisplay.BackBufferStride * BitmapToDisplay.PixelHeight ), ColorImageFormat.Bgra ); BitmapToDisplay.AddDirtyRect( new Int32Rect( 0, 0, _ColorFrame.FrameDescription.Width, _ColorFrame.FrameDescription.Height ) ); BitmapToDisplay.Unlock(); } With this snippet, I'm placing the converted Bgra32 color stream directly on the BackBuffer of the WriteableBitmap. This gives me pretty smooth playback, but I still get the occasional freeze for half a second. After a bit of profiling, I discovered there were a few problems. The first problem is the size of the buffer along with the conversion on the buffer. At this time, the raw image format of the data from the Kinect is Yuy2. This is great for direct video processing. It would be ideal if I had a WriteableVideo object in WPF. However, this is not the case. Further digging led me to the real problem. It appears that the SDK is converting the input serially. Let's think about this for a second. The color camera is a 1080p camera. As we should all know, this give us a native resolution of 1920 x 1080. This produces 2,073,600 pixels. Yuy2 uses 4 bytes per 2 pixel, for a buffer size of 4,147,200 bytes. Bgra32 uses 4 bytes per pixel, for a buffer size of 8,294,400 bytes. The SDK appears to be doing this on one thread. I started wondering if I chould do this better myself. I mean, I have 8 cores in my system. Why can't I use them all? The first problem is converting a Yuy2 frame into a Bgra32 frame. It is NOT trivial. I spent a day of research of just how to do this. In the end, I didn't even produce the best algorithm possible, but it did work. After I managed to get that to work, I knew my next step was the get the conversion operation off the UI Thread. This was a simple process of throwing the work into a Task. Of course, this meant I had to marshal the final write to the WriteableBitmap back to the UI thread. Finally, I needed to vectorize the operation so I could run it safely in parallel. This was, mercifully, not quite as hard as I thought it would be. I had my loop return an index to a pair of pixels. From there, I had to tell the loop to do everything for this pair of pixels. If you're wondering why I did it for pairs of pixels, look back above at the specification for the Yuy2 format. I won't go into full detail on why each 4 bytes contains 2 pixels of information, but rest assured that there is a reason why the format is described in that way. The first working attempt at this algorithm successfully turned my poor laptop into a space heater. I very quickly brought and maintained all 8 cores up to about 97% usage. That's when I remembered that obscure option in the Task Parallel Library where you could limit the amount of parallelism used. After a little trial and error, I discovered 4 parallel tasks was enough for most cases. This yielded the follow code: private byte ClipToByte( int p_ValueToClip ) { return Convert.ToByte( ( p_ValueToClip < byte.MinValue ) ? byte.MinValue : ( ( p_ValueToClip > byte.MaxValue ) ? byte.MaxValue : p_ValueToClip ) ); }   private void ColorFrameArrived( object sender, ColorFrameArrivedEventArgs e ) { if( null == e.FrameReference ) return;   // If you do not dispose of the frame, you never get another one... using( ColorFrame _ColorFrame = e.FrameReference.AcquireFrame() ) { if( null == _ColorFrame ) return;   byte[] _InputImage = new byte[_ColorFrame.FrameDescription.LengthInPixels * _ColorFrame.FrameDescription.BytesPerPixel]; byte[] _OutputImage = new byte[BitmapToDisplay.BackBufferStride * BitmapToDisplay.PixelHeight]; _ColorFrame.CopyRawFrameDataToArray( _InputImage );   Task.Factory.StartNew( () => { ParallelOptions _ParallelOptions = new ParallelOptions(); _ParallelOptions.MaxDegreeOfParallelism = 4;   Parallel.For( 0, Sensor.ColorFrameSource.FrameDescription.LengthInPixels / 2, _ParallelOptions, ( _Index ) => { // See http://msdn.microsoft.com/en-us/library/windows/desktop/dd206750(v=vs.85).aspx int _Y0 = _InputImage[( _Index << 2 ) + 0] - 16; int _U = _InputImage[( _Index << 2 ) + 1] - 128; int _Y1 = _InputImage[( _Index << 2 ) + 2] - 16; int _V = _InputImage[( _Index << 2 ) + 3] - 128;   byte _R = ClipToByte( ( 298 * _Y0 + 409 * _V + 128 ) >> 8 ); byte _G = ClipToByte( ( 298 * _Y0 - 100 * _U - 208 * _V + 128 ) >> 8 ); byte _B = ClipToByte( ( 298 * _Y0 + 516 * _U + 128 ) >> 8 );   _OutputImage[( _Index << 3 ) + 0] = _B; _OutputImage[( _Index << 3 ) + 1] = _G; _OutputImage[( _Index << 3 ) + 2] = _R; _OutputImage[( _Index << 3 ) + 3] = 0xFF; // A   _R = ClipToByte( ( 298 * _Y1 + 409 * _V + 128 ) >> 8 ); _G = ClipToByte( ( 298 * _Y1 - 100 * _U - 208 * _V + 128 ) >> 8 ); _B = ClipToByte( ( 298 * _Y1 + 516 * _U + 128 ) >> 8 );   _OutputImage[( _Index << 3 ) + 4] = _B; _OutputImage[( _Index << 3 ) + 5] = _G; _OutputImage[( _Index << 3 ) + 6] = _R; _OutputImage[( _Index << 3 ) + 7] = 0xFF; } );   Application.Current.Dispatcher.Invoke( () => { BitmapToDisplay.WritePixels( new Int32Rect( 0, 0, Sensor.ColorFrameSource.FrameDescription.Width, Sensor.ColorFrameSource.FrameDescription.Height ), _OutputImage, BitmapToDisplay.BackBufferStride, 0 ); } ); } ); } } This seemed to yield a results I wanted, but there was still the occasional stutter. This lead to what I realized was the second problem. There is a race condition between the UI Thread and me locking the WriteableBitmap so I can write the next frame. Again, I'm writing approximately 8MB to the back buffer. Then, I started thinking I could cheat. The Kinect is running at 30 frames per second. The WPF UI Thread runs at 60 frames per second. This made me not feel bad about exploiting the Composition Thread. I moved the bulk of the code from the FrameArrived handler into CompositionTarget.Rendering. Once I was in there, I polled from a frame, and rendered it if it existed. Since, in theory, I'm only killing the Composition Thread every other hit, I decided I was ok with this for cases where silky smooth video performance REALLY mattered. This ode looked like this: private byte ClipToByte( int p_ValueToClip ) { return Convert.ToByte( ( p_ValueToClip < byte.MinValue ) ? byte.MinValue : ( ( p_ValueToClip > byte.MaxValue ) ? byte.MaxValue : p_ValueToClip ) ); }   void CompositionTarget_Rendering( object sender, EventArgs e ) { using( ColorFrame _ColorFrame = FrameReader.AcquireLatestFrame() ) { if( null == _ColorFrame ) return;   byte[] _InputImage = new byte[_ColorFrame.FrameDescription.LengthInPixels * _ColorFrame.FrameDescription.BytesPerPixel]; byte[] _OutputImage = new byte[BitmapToDisplay.BackBufferStride * BitmapToDisplay.PixelHeight]; _ColorFrame.CopyRawFrameDataToArray( _InputImage );   ParallelOptions _ParallelOptions = new ParallelOptions(); _ParallelOptions.MaxDegreeOfParallelism = 4;   Parallel.For( 0, Sensor.ColorFrameSource.FrameDescription.LengthInPixels / 2, _ParallelOptions, ( _Index ) => { // See http://msdn.microsoft.com/en-us/library/windows/desktop/dd206750(v=vs.85).aspx int _Y0 = _InputImage[( _Index << 2 ) + 0] - 16; int _U = _InputImage[( _Index << 2 ) + 1] - 128; int _Y1 = _InputImage[( _Index << 2 ) + 2] - 16; int _V = _InputImage[( _Index << 2 ) + 3] - 128;   byte _R = ClipToByte( ( 298 * _Y0 + 409 * _V + 128 ) >> 8 ); byte _G = ClipToByte( ( 298 * _Y0 - 100 * _U - 208 * _V + 128 ) >> 8 ); byte _B = ClipToByte( ( 298 * _Y0 + 516 * _U + 128 ) >> 8 );   _OutputImage[( _Index << 3 ) + 0] = _B; _OutputImage[( _Index << 3 ) + 1] = _G; _OutputImage[( _Index << 3 ) + 2] = _R; _OutputImage[( _Index << 3 ) + 3] = 0xFF; // A   _R = ClipToByte( ( 298 * _Y1 + 409 * _V + 128 ) >> 8 ); _G = ClipToByte( ( 298 * _Y1 - 100 * _U - 208 * _V + 128 ) >> 8 ); _B = ClipToByte( ( 298 * _Y1 + 516 * _U + 128 ) >> 8 );   _OutputImage[( _Index << 3 ) + 4] = _B; _OutputImage[( _Index << 3 ) + 5] = _G; _OutputImage[( _Index << 3 ) + 6] = _R; _OutputImage[( _Index << 3 ) + 7] = 0xFF; } );   BitmapToDisplay.WritePixels( new Int32Rect( 0, 0, Sensor.ColorFrameSource.FrameDescription.Width, Sensor.ColorFrameSource.FrameDescription.Height ), _OutputImage, BitmapToDisplay.BackBufferStride, 0 ); } }

    Read the article

  • Oracle Customer Experience (CX) Solutions Make Retailers Merry

    - by Tuula Fai
    Tis the season to be jolly. If you’re a retailer, your level of jolliness depends on sales. So you watch trends like U.S. store traffic increasing 3.5% to 308 million on Black Friday but sales actually falling 1.8% to $11.2 billion. Fortunately, by the end of November, retail sales were up 3.7% over the previous year, thanks to life recovering after Hurricane Sandy. And online sales topped $1 billion for the first time ever! Who are the companies improving their sales online? They are big names like Walgreen’s Drugstore.com, Nordstrom’s HauteLook, and Intuit. More importantly, how are they doing it? They use cutting-edge business practices enabled by Oracle’s CX Cloud Service & Support solutions to: Increase conversions rates and order sizes (Customer Acquisition) Enhance customer satisfaction and loyalty (Customer Retention) Reduce contact center costs and improve agent productivity (Operational Efficiency). Acquisition + Retention + Operational Efficiency = Sustainable Growth and Profits. That’s the magic formula for retail customer service success. Don’t take our word for it. Look at the results of these Oracle customers: Walgreen’s Drugstore—30% sales conversion rate on chat sessions with 20% increase in shopping cart size Nordstrom’s HauteLook—40,000+ interactions per month—20% growth over last year— efficiently managed by 40 agents, with no increase in IT costs Intuit—50% increase in customer satisfaction and 70% decrease in cost per interaction Using Oracle’s CX Cloud & Service solutions, these retailers deliver consistent, relevant, and personalized experiences across all touchpoints, including social, mobile, and web. Their ability to connect with customers anytime, anywhere—providing the right answer at the right time—helps them create a defensible advantage in the marketplace. Want to learn more? Please visit http://www.oracle.com/goto/cloudlaunchpad for free resources on delivering exceptional customer service in the Cloud. Also, watch our YouTube channel to learn more about seamless multichannel retail and Winston Furnishings’ exceptional customer experience.

    Read the article

  • The clock problem - to if or not to if?

    - by trejder
    Let's say, we have a simple digital clock. To "power" it, we use a routine executed every second. We update seconds part in it. But, what about minutes and hours part? What is better / more professional / offers better performance: Ignore all checking and update hour, minute and seconds part each time, every second. Use if + a variable for checking, if 60 (or 3600) seconds passed and update minute / hour part only at that precise moments. This leads us to a question, what is better -- unnecessary drawings (first approach) or extra ifs? I've just spotted a Javascript digital clock, one of millions similar on one of billions pages. And I noticed that all three parts (hours, minutes and seconds) are updated every second, though first changes its value only once per 3600 seconds and second once per 60 seconds. I'm not to experienced developer, so I might me wrong. But everything, what I've learnt up until now, tells me, that if are far better then executing drawing / refreshing sequences only to draw the same content.

    Read the article

  • Save Actions in NetBeans IDE 7.3

    - by Geertjan
    Several developers, especially those familiar with equivalent functionality in Eclipse, have been asking for so-called "Save Actions", that is, support for actions that are automatically performed when a file is saved. Here's the related NetBeans issue: http://netbeans.org/bugzilla/show_bug.cgi?id=140719   In NetBeans IDE 7.3, the issue is resolved as follows: A new "On Save" tab is found in the "Editor" tab of the Options window. Defaults for all languages are set via the "All Languages" item in the drop-down. Here, for all languages, you can specify what kind (all, none, or only modified lines) of formatting and space removal will occur automatically when a file is saved: Via the drop-down, you see all the languages supported by the IDE: You can pick a language and then override the default On Save settings: Per language, there may be additional On Save settings. For example, for Java, you can specify that, when saving a Java file, unused import statements should be removed and/or the rules you've set for organizing import statements should be applied: There's also a set of new NetBeans IDE APIs for adding new On Save functionality via custom plugins. Via MIME type registration of OnSaveTask.Factory, you can register new On Save actions that will be run for files conforming to the relevant MIME type. There's also extensions via the Editor Options API for registering new panels (one per language) to the On Save panel in the Options window. I'll demonstrate some examples of the APIs in upcoming blog entries.

    Read the article

  • Dealing with Fine-Grained Cache Entries in Coherence

    - by jpurdy
    On occasion we have seen significant memory overhead when using very small cache entries. Consider the case where there is a small key (say a synthetic key stored in a long) and a small value (perhaps a number or short string). With most backing maps, each cache entry will require an instance of Map.Entry, and in the case of a LocalCache backing map (used for expiry and eviction), there is additional metadata stored (such as last access time). Given the size of this data (usually a few dozen bytes) and the granularity of Java memory allocation (often a minimum of 32 bytes per object, depending on the specific JVM implementation), it is easily possible to end up with the case where the cache entry appears to be a couple dozen bytes but ends up occupying several hundred bytes of actual heap, resulting in anywhere from a 5x to 10x increase in stated memory requirements. In most cases, this increase applies to only a few small NamedCaches, and is inconsequential -- but in some cases it might apply to one or more very large NamedCaches, in which case it may dominate memory sizing calculations. Ultimately, the requirement is to avoid the per-entry overhead, which can be done either at the application level by grouping multiple logical entries into single cache entries, or at the backing map level, again by combining multiple entries into a smaller number of larger heap objects. At the application level, it may be possible to combine objects based on parent-child or sibling relationships (basically the same requirements that would apply to using partition affinity). If there is no natural relationship, it may still be possible to combine objects, effectively using a Coherence NamedCache as a "map of maps". This forces the application to first find a collection of objects (by performing a partial hash) and then to look within that collection for the desired object. This is most naturally implemented as a collection of entry processors to avoid pulling unnecessary data back to the client (and also to encapsulate that logic within a service layer). At the backing map level, the NIO storage option keeps keys on heap, and so has limited benefit for this situation. The Elastic Data features of Coherence naturally combine entries into larger heap objects, with the caveat that only data -- and not indexes -- can be stored in Elastic Data.

    Read the article

  • International Pricing of Software [closed]

    - by arachnode.net
    I operate a small company that charges $99 for a piece of software. I'd like to know what would be a fair price for non-US customers. Today I sold a license to a party in South Africa. He told me he had been watching the project for two years while business justification could be made for the purchase as SA's currency is nine times weaker than the US dollar. I found this resource detailing how much a Big Mac costs in various countries: http://howmuchatyourplace.com/how_much_does/Big%20Mac_cost.php I realize that the cost of producing a Big Mac varies from locale to locale as does the demand for one. I am aware that many software companies charge prices in local currencies that equate to the price in US dollars. I am aware that my costs remain fixed, and I obviously I cannot discount the rate at which my time costs me. I'm OK with earning less per sale as I would rather get my software onto the desktops of those that need it rather than having them try to write it themselves. Support is light and I can usually point a user to an existing blog or forum post. Being a resident of Hawaii, I am aware that certain goods and services cost more here. Power is up to six times as much per KWH as it is in, say, Seattle, and wages are approximately 60% of what they are for my profession (programmer). I'd like to offer my software at a price that would be fair for everyone around the globe. If a currency is 2 foreign units to 1 US dollar, and goods and services cost 50% more and pay for an equivalent job is 50% of what it is here, should I charge, say, $50 instead of $99? Is there a resource which would allow me to input a price in US dollars and adjust for a list of international locations?

    Read the article

  • Trade In, Trade Up Promotion: SPARC Consolidation Now Through May 31st

    - by swalker
    Dear Partner, Installed Base Business (IBB) technology refresh is one of the most important activities for Oracle, for you and for your customers. It allows your existing customers to benefit from the most up-to-date, best-of-breed Oracle products. And it’s an exciting time to perform a technology refresh: a new SPARC promotion is available now, closing 31st May 2012. Customers trading in older SPARC systems and upgrading to a new SPARC SuperCluster T4-4 or SPARC Enterprise M8000/M9000 can get $4,000 per CPU. Discount is pre-approved and upfront (maximum discounts apply). The major highlights are as follows: Targeted Systems: Upgrade to SPARC M8000, M9000, SuperCluster Qualified installed base upgrade from: All older-generations of SPARC systemsPromotional offer: Trade-in Value: $4K per CPU Pre-approved maximum discount (including trade-in) not to exceed 60% on M8/9000 systems and 25% on SuperCluster No-cost dock-to-dock shipping, and environmentally safe disposal of the returned hardware through Oracle best-of-class recycling processes. Recommendations: We recommend you to take the following actions: As usual, please register your opportunities in OMM When you do so, please make sure you place the following Campaign Names in the “Marketing Initiative” field of OMM: Campaign Name : EMEA_Tech Refresh-IBB Campaign_12H1_Follow Up_O For all the details: Please view rules, and FAQs. For more information, please visit the Promo Partner Site here. For more information on IBB and the Oracle Upgrade Advantage Program (UAP):http://www.oracle.com/us/products/servers-storage/upgrade-advantage-program/index.html http://www.oracle.com/partners/secure/sales/oracle-ibb-program-for-partners-184291.html Contacts: For questions, please contact your favorite Oracle Partner Account Manager.

    Read the article

  • What are solutions and tradeoffs to maintain search result consistency in a web application

    - by iammichael
    Consider a web application with a custom search function that must display the results in a paged manner (twenty per page with up to hundreds of thousands of total results) and the ability to drill down to individual results that maintain next/previous links to navigate through the results. Re-executing the search on each page request to get the appropriate results for that page of data can be too expensive (up to 15s per search). Also, since the underlying data can change frequently (e.g. addition of new results), re-executing could cause the next/previous functionality to result in inconsistent behavior (e.g. the same results reappearing on a later page after having been viewed on an earlier page). What options exist to ensure the search results can be viewed across multiple pages in a consistent manner, and what tradeoffs does each option have in terms of network, CPU, memory, and storage requirements? EDIT: I thought caching the query search results was an obvious necessity. The question is really asking about where to cache the result set and what tradeoffs might exist to each. For example, storing the ids of the entities in the result set on the client, or storing the IDs of the entities themselves in the users session on the web server, or in a temporary table in the database. I'm not looking specifically for a single solution as different scenarios may result in different approaches (and such a question would be more suited for stackoverflow.com rather than here), but more of a design comparison between the possible approaches.

    Read the article

  • Upgraded from 11.4 to 11.10, There was an error, now the system won't initiialize

    - by Eric
    This morning the system gave me a message that my Version (11.4) was no longer supported, and I took the 'upgrade' option (- 11.10). While installing the various components I encountered a message to the effect that the there was an error and the system may have become unusable. Among the messges: E:Sup-process /usr/bin/dpkg received a segmentation fault...returned an error code (1). I was given an option to do several things, one of which seemed to mean that it would attempt to roll back to the previous version (the default), which I took. After the process ran it said the upgrade process had finished, but there were errors. I attempted to initialize a console so I could enter ubuntu-bug update-manager /var/log/dist-upgrade, per the instructions I received when I received the error message, but the console failed during initialization. I restarted the machine, and the screen has stopped with the following contents: * Starting bluetooth * Stopping save kernel messages * Starting CUPS printing spooler/server * PulseAudio configured per-user sessions saned disabled: edit /etc/default/saned $starting up Cisco VPN daemon *Starting anac(h)ronistic cron *Stopping anac(h)ronistic cron Each of these steps followed by [ OK ] What are my options? Any help appreciated!

    Read the article

  • How much a programmer should read in order to keep himself updated? [closed]

    - by anything
    There are lots of technical books available. Below are few links which lists some good books If you could only have one programming related book on your bookshelf what would it be and why? What non-programming books should a programmer read to help develop programming/thinking skills? Best books on the theory and practice of software architecture? http://stackoverflow.com/questions/1711/what-is-the-single-most-influential-book-every-programmer-should-read ... and the list can go on and on and on. It will be really difficult to read all of the above mentioned books. I am not sure if its even possible for anyone to do that. Even if you filter it based on one's area of interest or work, list is still very large. .. and the technology keeps on changing (even more books :-( ) So, my question is how much a programmer should read lets say per year? How much hours one should put in such activities to keep oneself up to date? How do we find out the time required? PS: Average programmer reads less than one book per year (Code complete). What about the good programmers?

    Read the article

  • What should a programmer's yearly routine be to maximize their technical skills?

    - by sguptaet
    2 years ago I made a big career change into programming. I learned various technologies on my own without any prior experience. I really love it and feel lucky with all the resources around us to help us learn. Books, courses, open-source, etc. There are so many avenues. I'm wondering what a good routine would be to follow to maximize my software development skills. I don't believe just building software is the way, because that leaves no time for learning new concepts or technologies. I'm looking for an answer like this: Take a new concept sabbatical/workshop 2 weeks per year. Read 1 theoretical and 1 practical programming book per year. Learn 1 additional language every 2 years. Take a 1 week vacation every 6 months. Etc. I realize that the above might sound naive and unrealistic as there are so many factors. But I'd like to know the "recipe" that you think is best that will serve as a guide for people.

    Read the article

  • Helping to Reduce Page Compression Failures Rate

    - by Vasil Dimov
    When InnoDB compresses a page it needs the result to fit into its predetermined compressed page size (specified with KEY_BLOCK_SIZE). When the result does not fit we call that a compression failure. In this case InnoDB needs to split up the page and try to compress again. That said, compression failures are bad for performance and should be minimized.Whether the result of the compression will fit largely depends on the data being compressed and some tables and/or indexes may contain more compressible data than others. And so it would be nice if the compression failure rate, along with other compression stats, could be monitored on a per table or even on a per index basis, wouldn't it?This is where the new INFORMATION_SCHEMA table in MySQL 5.6 kicks in. INFORMATION_SCHEMA.INNODB_CMP_PER_INDEX provides exactly this helpful information. It contains the following fields: +-----------------+--------------+------+ | Field | Type | Null | +-----------------+--------------+------+ | database_name | varchar(192) | NO | | table_name | varchar(192) | NO | | index_name | varchar(192) | NO | | compress_ops | int(11) | NO | | compress_ops_ok | int(11) | NO | | compress_time | int(11) | NO | | uncompress_ops | int(11) | NO | | uncompress_time | int(11) | NO | +-----------------+--------------+------+ similarly to INFORMATION_SCHEMA.INNODB_CMP, but this time the data is grouped by "database_name,table_name,index_name" instead of by "page_size".So a query like SELECT database_name, table_name, index_name, compress_ops - compress_ops_ok AS failures FROM information_schema.innodb_cmp_per_index ORDER BY failures DESC; would reveal the most problematic tables and indexes that have the highest compression failure rate.From there on the way to improving performance would be to try to increase the compressed page size or change the structure of the table/indexes or the data being stored and see if it will have a positive impact on performance.

    Read the article

  • HTML5 - Does it have the power to handle a large 2D game with a huge world?

    - by user15858
    I have been using XNA game studio, but due to private reasons (as well as the ability to publish anywhere & my heavy interest in isogenic engine), I would like to switch to HTML5. However, I have very high 2D graphic demands for my game. The game itself will have a HDD size of anywhere between 6GB (min) to 12GB (max) which would be a full game deployed offline. The size of the images aren't significantly large, so streaming would be entirely possible if only those assets required were streamed as needed. The game has a massive file size because of the sheer amount of content. For some images or spritesheets, they would be quite massive. (ex. a very large Dragon, which if animated in a spritesheet would be split into two 4096x4096 sheets or one 8192x8192 sheet). Most assets would be very small, and about 7MB for a full character with 15 animations in every direction (all animations not required immediately) so in the size of a few hundred KB to download before the game loads. My question, however, is if the graphical power of HTML5 is enough to animate several characters on screen at once, when it flips through frames quite rapidly. All my sprites have about 25 frames per animation, 5 directions (a spritesheet for each direction & animation), and run at 30fps. Upon changing direction, animation, or a new character entering, spritesheets would change and be constantly loading/unloading. If I pack all directions in a single sheet, it would be about 2048x2048 per sheet. Most frameworks have no problem with this, but I am afraid from what I read that HTML5's graphical capabilities will limit me. Since it takes significant time simply to animate characters in any language, I'd like a quick answer.

    Read the article

  • Would this data requirement suit a Document -Oriented database?

    - by codecowboy
    I have a requirement to allow users to fill in journal/diary entries per day. I want to provide a handful of known journal templates with x columns to fill in. An example might be a thought diary; a user has to record a thought in one column, describe the situation, rate how they felt etc. The other requirement is that a user should be able to create their own diary templates. They might have a need for a 10 column diary entry per day and might need to rate some aspect out of 50 instead of 10. In an RDBMS, I can see this getting quite complicated. I could have individual tables for my known templates as the fields will be fixed. But for custom diary templates I imagine I would would need a table storing custom_field_types (the diary columns), a table storing entries referencing their field types (custom_entries) and then a third custom_diary table which would store rows matching custom_entries to diaries. Leaving performance / scaling aside, would it be any simpler or make more sense to use a document oriented database like MongoDB to store this data? This is for a web application which might later need an API for mobile devices.

    Read the article

  • What's the best way to use requestAnimationFrame and fixed frame rates

    - by m90
    I recently got into using the HTML5-requestAnimationFrame-API a lot on animation-heavy websites, especially after seeing the Jank Busters talk. This seems to work pretty well and really improve performance in many cases. Yet one question still persists for me: When wanting to use an animation that is NOT entirely calculated (think spritesheets for example) you will have to aim for a fixed frame rate. Of course one could go back to use setInterval again, but maybe there are other ways to tackle this. The two ways I could think of using requestAnimationFrame with a fixed frame rate are: var fps = 25; //frames per second function animate(){ //actual drawing goes here setTimeout(function(){ requestAnimationFrame(animate); }, 1000 / fps) } animate(); or var fps = 25; //frames per second var lastExecution = new Date().getTime(); function animate(){ var now = new Date().getTime(); if ((now - lastExecution) > (1000 / fps)){ //do actual drawing lastExecution = new Date().getTime(); } requestAnimationFrame(animate); } animate(); Personally, I'd opt for the second option (the first one feels like cheating), yet it seems to be more buggy in certain situations. Is this approach really worth it (especially at low frame rates like 12.5)? Are there things to be improved? Is there another way to tackle this?

    Read the article

  • Simple (and fast) dices physics

    - by Markus von Broady
    I'm programming a throw of 5 dices in Actionscript 3 + AwayPhysics (BulletPhysics port). I had a lot of fun tweaking frictions, masses etc. and in the end I found best results with more physics ticks per frame. Currently I use 10 ticks per frame (1/60 s) and it's OK, though I see a difference in plus for 20 ticks. Even though it's only 5 cubes (dices) in a box (or a floor with 3 walls really) I can't simulate 20 ticks in a frame and keep FPS at 60 on a medium-aged PC. That's why I decided to precompute frames for animation, finishing it in around 1700 ticks in 2 seconds. The flash player is freezed for these 2 seconds, and I'm afraid that this result will be more of a 5 seconds or even more, if I'll simulate multi-threading and compute frames in background of some other heavy processes and CPU drawing (dices is only a part of this game). Because I want both players to see dices roll in same way, I can't compute physics when having free resources, and build a buffer for at least one throw of each type (where type is number of dices thrown). I'm afraid players will see a "preparing dices........." message too often and for too long. I think the only solution to this problem is replacing PhysicsEngine with something simpler, or creating own physicsEngine. Do You have any formulas for cube-cube and cube-wall collision detection, and for calculating how their angular and linear velocities should change after a collision occurs?

    Read the article

  • Need some critique on .NET/WCF SOA architecture plan

    - by user998101
    I am working on a refactoring of some services and would appreciate some critique on my general approach. I am working with three back-end data systems and need to expose an authenticated front-end API over http binding, JSON, and REST for internal apps as well as 3rd party integration. I've got a rough idea below that's a hybrid of what I have and where I intend to wind up. I intend to build guidance extensions to support this architecture so that devs can build this out quickly. Here's the current idea for our structure: Front-end WCF routing service (spread across multiple IIS servers via hardware load balancer) Load balancing of services behind routing is handled within routing service, probably round-robin One of the services will be a token Multiple bindings per-service exposed to address JSON, REST, and whatever else comes up later All in/out is handled via POCO DTOs Use unity to scan for what services are available and expose them The front-end services behind the routing service do nothing more than expose the API and do conversion of DTO<-Entity Unity inject service implementation to allow mocking automapper for DTO/Entity conversion Invoke WF services where response required immediately Queue to ESB for async WF -- ESB will invoke WF later Business logic WF layer Expose same api as front-end services Implement business logic Wrap transaction context where needed Call out to composite/atomic services Composite/Atomic Services Exposed as WCF One service per back-end system Standard atomic CRUD operations plus composite operations Supports transaction context The questions I have are: Are the separation of concerns outlined above beneficial? Current thought is each layer below is its own project, except the backend stuff, where each system gets one project. The project has a servicehost and all the services are under a services folder. Interfaces live in a separate project at each layer. DTO and Entities are in two separate projects under a shared folder. I am currently planning to build dedicated services for shared functionality such as logging and overload things like tracelistener to call those services. Is this a valid approach? Any other suggestions/comments?

    Read the article

  • What is the best way to manage large 3d worlds (i.e minecraft style)?

    - by SomeXnaChump
    After playing minecraft I was marvelling a bit at their large worlds but at the same time finding it extremely slow to navigate, even with a quad core and meaty graphics card. Now I assume its fairly slow because: A) Its written in Java, and as most of the actual spatial partitioning and other memory management activities happen in there it would be slower than a native C++ version. B) They are not partitioning their world very well I could be wrong on both assumptions, however it got me thinking about the best way to manage large worlds. As it is more of a true 3d world, where a block can exist in any part of the world, it is basically a big 3d array [x][y][z], where each block in the world has a type (i.e BlockType.Empty = 0, BlockType.Dirt = 1 etc). Now I am assuming to make this sort of world performant you would need to: a) Use a tree of some variety (oct/kd/bsp) to split all the cubes out, it seems like an oct/kd would be the better option as you can just partition on a per cube level not a per triangle level. b) Use some algorithm to work out if the blocks within the scene can currently be seen, as blocks closer to the user could obfuscate the blocks behind, making it pointless to render them. c) Keep the block object themselves lightweight, so it is quick to add and remove them from the trees I guess there is no right answer to this, but I would be interested to see peoples opinions on the subject.

    Read the article

  • Internal Libraries (Subversion Externals, 'library' branch, or just another folder)

    - by Ntsc
    Currently working on multiple projects that need to share internal libraries. The internal libraries are updated continually. Currently only 1 project needs to be stable but soon we will need to have both projects stable at any given time. What is the best way to SVN internal libraries? Currently we are using the 'just another folder' like so... trunk\project1 trunk\project2 trunk\libs It causes a major headache when a shared library is updated for project1 and project2 is now dead until the parts that use the library are updated. So after doing some research on SVN externals I thought of this... trunk\project1\libs (external to trunk\libs @ some revision) trunk\project2\libs (external to trunk\libs @ different revision) trunk\libs\ I'm a little worried about how externals work with commits and not making library commits so complicated that I am the only one capable of doing it (mostly worried about branches with externals as we use them extensively). On top of that we have multiple programming languages within each project some of which don't support per-project library directories (at least not easily) so we would need to check out on a per project basis instead of checking out the trunk. There is also the 'vendor' style branching of libraries but it has the same problem as above where the library would have to be a sub folder of each project and is maybe a little to complicated for how little projects we have. Any insight would be nice. I've spent quite a bit of time reading the Subversion book and feeling like I'm getting no where.

    Read the article

  • Survey Data Model - How to avoid EAV and excessive denormalization?

    - by AlexDPC
    Hi everyone, My database skills are mediocre at best and I have to design a data model for survey data. I have spent some thoughts on this and right now I feel that I am stuck between some kind of EAV model and a design involving hundreds of tables, each with hundreds of columns (and thousands of records). There must be a better way to do this and I hope that the wise folks on this forum can help me. I have already searched various forums, but I couldn't really find a solution. If it has already been given elsewhere, please excuse me and provide me with a link so I can read it up. Some assumptions about the data I have to deal with: Each survey consists of 1 to n questionnaires Each questionnaire consists of 100-2,000 questions (please ignore that 2,000 questions really sound like a lot to answer...) Questions can be of various types: multiple-choice, free text, a number (like age, income, percentages, ...) Each survey involves 10-200 countries (These are not the respondents. The respondents are actually people in the countries.) Depending on the type of questionnaire, each questionnaire is answered by 100-20,000 respondents per country. A country can adapt the questionnaires for a survey, i.e. add, remove or edit questions The data for one country is gathered in a separate database in that country. There is no possibility for online integration from the start. The data for all countries has to be integrated later. This means for example, if a country has deleted a question, that data must somehow be derived from what they sent in order to achieve a uniform design across all countries I will have to write the integration and cleaning software, which will need to work with every country's data In the end the data needs to be exported to flat files, one rectangular grid per country and questionnaire. I have already discussed this topic with people from various backgrounds and have not come to a good solution yet. I mainly got two kinds of opinions. The domain experts, who are used to working with flat files (spreadsheet-style) for data processing and analysis vote for a denormalized structure with loads of tables and columns as I described above (1 table per country and questionnaire). This sounds terrible to me, because I learned that wide tables are to be avoided, it will be annoying to determine which columns are actually in a table when working with it, the database will become cluttered with hundreds of tables (or I even need to set up multiple databases, each with a similar yet a bit differetn design), etc. O-O-programmers vote for a strongly "normalized" design, which would effectively lead to a central table containing all the answers from all respondents to all questions. This table would either need to contain a column of type sql_variant type or multiple answer columns with different types to store answers of different types (multiple choice, free text, ..). The former would essentially be a EAV model. I tend to follow Joe Celko here, who strongly discourages its use (he calls it OTLT or "One True Lookup Table"). The latter would imply that each row would contain null cells for the not applicable types by design. Another alternative I could think of would be to create one table per answer type, i.e., one for multiple-choice questions, one for free text questions, etc.. That's not so generic, it would lead to a lot of union joins, I think and I would have to add a table if a new answer type is invented. Sorry for boring you with all this text and thank you for your input! Cheers, Alex PS: I asked the same question here: http://www.eggheadcafe.com/community/aspnet/13/10242616/survey-data-model--how-to-avoid-eav-and-excessive-denormalization.aspx

    Read the article

  • SQL - date variable isn't being parsed correctly?

    - by Bill Sambrone
    I am pulling a list of invoices filtered by a starting and ending date, and further filtered by type of invoice from a SQL table. When I specify a range of 2013-07-01 through 2013-09-30 I am receiving 2 invoices per company when I expect 3. When I use the built in select top 1000 query in SSMS and add my date filters, all the expected invoices appear. Here is my fancy query that I'm using that utilizing variables that are fed in: DECLARE @ReportStart datetime DECLARE @ReportStop datetime SET @ReportStart = '2013-07-01' SET @ReportStop = '2013-09-30' SELECT Entity_Company.CompanyName, Reporting_AgreementTypes.Description, Reporting_Invoices.InvoiceAmount, ISNULL(Reporting_ProductCost.ProductCost,0), (Reporting_Invoices.InvoiceAmount - ISNULL(Reporting_ProductCost.ProductCost,0)), (Reporting_AgreementTypes.Description + Entity_Company.CompanyName), Reporting_Invoices.InvoiceDate FROM Reporting_Invoices JOIN Entity_Company ON Entity_Company.ClientID = Reporting_Invoices.ClientID LEFT JOIN Reporting_ProductCost ON Reporting_ProductCost.InvoiceNumber =Reporting_Invoices.InvoiceNumber JOIN Reporting_AgreementTypes ON Reporting_AgreementTypes.AgreementTypeID = Reporting_Invoices.AgreementTypeID WHERE Reporting_Invoices.AgreementTypeID = (SELECT AgreementTypeID FROM Reporting_AgreementTypes WHERE Description = 'Resold Services') AND Reporting_Invoices.InvoiceDate >= @ReportStart AND Reporting_Invoices.InvoiceDate <= @ReportStop ORDER BY CompanyName,InvoiceDate The above only returns 2 invoices per company. When I run a much more basic query through SSMS I get 3 as expected, which looks like: SELECT TOP 1000 [InvoiceID] ,[AgreementID] ,[AgreementTypeID] ,[InvoiceDate] ,[Comment] ,[InvoiceAmount] ,[InvoiceNumber] ,[TicketID] ,Entity_Company.CompanyName FROM Reporting_Invoices JOIN Entity_Company ON Entity_Company.ClientID = Reporting_Invoices.ClientID WHERE Entity_Company.ClientID = '9' AND AgreementTypeID = (SELECT AgreementTypeID FROM Reporting_AgreementTypes WHERE Description = 'Resold Services') AND Reporting_Invoices.InvoiceDate >= '2013-07-01' AND Reporting_Invoices.InvoiceDate <= '2013-09-30' ORDER BY InvoiceDate DESC I've tried stripping down the 1st query to include only a client ID on the original invoice table, the invoice date, and nothing else. Still only get 2 invoices instead of the expected 3. I've also tried manually entering the dates instead of the @ variables, same result. I confirmed that InvoiceDate is defined as a datetime in the table. I've tried making all JOIN's a FULL JOIN to see if anything is hiding, but no change. Here is how I stripped down the original query to keep all other tables out of the mix and yet I'm still getting only 2 invoices per client ID instead of 3 (I manually entered the ID for the type filter): --DECLARE @ReportStart datetime --DECLARE @ReportStop datetime --SET @ReportStart = '2013-07-01' --SET @ReportStop = '2013-09-30' SELECT --Entity_Company.CompanyName, --Reporting_AgreementTypes.Description, Reporting_Invoices.ClientID, Reporting_Invoices.InvoiceAmount, --ISNULL(Reporting_ProductCost.ProductCost,0), --(Reporting_Invoices.InvoiceAmount - ISNULL(Reporting_ProductCost.ProductCost,0)), --(Reporting_AgreementTypes.Description + Entity_Company.CompanyName), Reporting_Invoices.InvoiceDate FROM Reporting_Invoices --JOIN Entity_Company ON Entity_Company.ClientID = Reporting_Invoices.ClientID --LEFT JOIN Reporting_ProductCost ON Reporting_ProductCost.InvoiceNumber = Reporting_Invoices.InvoiceNumber --JOIN Reporting_AgreementTypes ON Reporting_AgreementTypes.AgreementTypeID = Reporting_Invoices.AgreementTypeID WHERE Reporting_Invoices.AgreementTypeID = '22'-- (SELECT AgreementTypeID FROM Reporting_AgreementTypes WHERE Description = 'Resold Services') AND Reporting_Invoices.InvoiceDate >= '2013-07-01' AND Reporting_Invoices.InvoiceDate <= '2013-09-30' ORDER BY ClientID,InvoiceDate This strikes me as really weird as it is pretty much the same query as the SSMS generated one that returns correct results. What am I overlooking? UPDATE I've further refined my "test query" that is returning only 2 invoices per company to help troubleshoot this. Below is the query and a relevant subset of data for 1 company from the appropriate tables: SELECT Reporting_Invoices.ClientID, Reporting_AgreementTypes.Description, Reporting_Invoices.InvoiceAmount, Reporting_Invoices.InvoiceDate FROM Reporting_Invoices JOIN Reporting_AgreementTypes ON Reporting_AgreementTypes.AgreementTypeID = Reporting_Invoices.AgreementTypeID WHERE Reporting_Invoices.AgreementTypeID = (SELECT AgreementTypeID FROM Reporting_AgreementTypes WHERE Description = 'Resold Services') AND Reporting_Invoices.InvoiceDate >= '2013-07-01T00:00:00' AND Reporting_Invoices.InvoiceDate <= '2013-09-30T00:00:00' ORDER BY Reporting_Invoices.ClientID,InvoiceDate The above only returns 2 invoices. Here is the relevant table data: Relevant data from Reporting_AgreementTypes AgreementTypeID Description 22 Resold Services Relevant data from Reporting_Invoices InvoiceID ClientID AgreementID AgreementTypeID InvoiceDate 16111 9 757 22 2013-09-30 00:00:00.000 15790 9 757 22 2013-08-30 00:00:00.000 15517 9 757 22 2013-07-31 00:00:00.000 Actual results from my new modified query ClientID Description InvoiceAmount InvoiceDate 9 Resold Services 3513.79 7/31/13 00:00:00 9 Resold Services 3570.49 8/30/13 00:00:00

    Read the article

  • What is the fastest cyclic synchronization in Java (ExecutorService vs. CyclicBarrier vs. X)?

    - by Alex Dunlop
    Which Java synchronization construct is likely to provide the best performance for a concurrent, iterative processing scenario with a fixed number of threads like the one outlined below? After experimenting on my own for a while (using ExecutorService and CyclicBarrier) and being somewhat surprised by the results, I would be grateful for some expert advice and maybe some new ideas. Existing questions here do not seem to focus primarily on performance, hence this new one. Thanks in advance! The core of the app is a simple iterative data processing algorithm, parallelized to the spread the computational load across 8 cores on a Mac Pro, running OS X 10.6 and Java 1.6.0_07. The data to be processed is split into 8 blocks and each block is fed to a Runnable to be executed by one of a fixed number of threads. Parallelizing the algorithm was fairly straightforward, and it functionally works as desired, but its performance is not yet what I think it could be. The app seems to spend a lot of time in system calls synchronizing, so after some profiling I wonder whether I selected the most appropriate synchronization mechanism(s). A key requirement of the algorithm is that it needs to proceed in stages, so the threads need to sync up at the end of each stage. The main thread prepares the work (very low overhead), passes it to the threads, lets them work on it, then proceeds when all threads are done, rearranges the work (again very low overhead) and repeats the cycle. The machine is dedicated to this task, Garbage Collection is minimized by using per-thread pools of pre-allocated items, and the number of threads can be fixed (no incoming requests or the like, just one thread per CPU core). V1 - ExecutorService My first implementation used an ExecutorService with 8 worker threads. The program creates 8 tasks holding the work and then lets them work on it, roughly like this: // create one thread per CPU executorService = Executors.newFixedThreadPool( 8 ); ... // now process data in cycles while( ...) { // package data into 8 work items ... // create one Callable task per work item ... // submit the Callables to the worker threads executorService.invokeAll( taskList ); } This works well functionally (it does what it should), and for very large work items indeed all 8 CPUs become highly loaded, as much as the processing algorithm would be expected to allow (some work items will finish faster than others, then idle). However, as the work items become smaller (and this is not really under the program's control), the user CPU load shrinks dramatically: blocksize | system | user | cycles/sec 256k 1.8% 85% 1.30 64k 2.5% 77% 5.6 16k 4% 64% 22.5 4096 8% 56% 86 1024 13% 38% 227 256 17% 19% 420 64 19% 17% 948 16 19% 13% 1626 Legend: - block size = size of the work item (= computational steps) - system = system load, as shown in OS X Activity Monitor (red bar) - user = user load, as shown in OS X Activity Monitor (green bar) - cycles/sec = iterations through the main while loop, more is better The primary area of concern here is the high percentage of time spent in the system, which appears to be driven by thread synchronization calls. As expected, for smaller work items, ExecutorService.invokeAll() will require relatively more effort to sync up the threads versus the amount of work being performed in each thread. But since ExecutorService is more generic than it would need to be for this use case (it can queue tasks for threads if there are more tasks than cores), I though maybe there would be a leaner synchronization construct. V2 - CyclicBarrier The next implementation used a CyclicBarrier to sync up the threads before receiving work and after completing it, roughly as follows: main() { // create the barrier barrier = new CyclicBarrier( 8 + 1 ); // create Runable for thread, tell it about the barrier Runnable task = new WorkerThreadRunnable( barrier ); // start the threads for( int i = 0; i < 8; i++ ) { // create one thread per core new Thread( task ).start(); } while( ... ) { // tell threads about the work ... // N threads + this will call await(), then system proceeds barrier.await(); // ... now worker threads work on the work... // wait for worker threads to finish barrier.await(); } } class WorkerThreadRunnable implements Runnable { CyclicBarrier barrier; WorkerThreadRunnable( CyclicBarrier barrier ) { this.barrier = barrier; } public void run() { while( true ) { // wait for work barrier.await(); // do the work ... // wait for everyone else to finish barrier.await(); } } } Again, this works well functionally (it does what it should), and for very large work items indeed all 8 CPUs become highly loaded, as before. However, as the work items become smaller, the load still shrinks dramatically: blocksize | system | user | cycles/sec 256k 1.9% 85% 1.30 64k 2.7% 78% 6.1 16k 5.5% 52% 25 4096 9% 29% 64 1024 11% 15% 117 256 12% 8% 169 64 12% 6.5% 285 16 12% 6% 377 For large work items, synchronization is negligible and the performance is identical to V1. But unexpectedly, the results of the (highly specialized) CyclicBarrier seem MUCH WORSE than those for the (generic) ExecutorService: throughput (cycles/sec) is only about 1/4th of V1. A preliminary conclusion would be that even though this seems to be the advertised ideal use case for CyclicBarrier, it performs much worse than the generic ExecutorService. V3 - Wait/Notify + CyclicBarrier It seemed worth a try to replace the first cyclic barrier await() with a simple wait/notify mechanism: main() { // create the barrier // create Runable for thread, tell it about the barrier // start the threads while( ... ) { // tell threads about the work // for each: workerThreadRunnable.setWorkItem( ... ); // ... now worker threads work on the work... // wait for worker threads to finish barrier.await(); } } class WorkerThreadRunnable implements Runnable { CyclicBarrier barrier; @NotNull volatile private Callable<Integer> workItem; WorkerThreadRunnable( CyclicBarrier barrier ) { this.barrier = barrier; this.workItem = NO_WORK; } final protected void setWorkItem( @NotNull final Callable<Integer> callable ) { synchronized( this ) { workItem = callable; notify(); } } public void run() { while( true ) { // wait for work while( true ) { synchronized( this ) { if( workItem != NO_WORK ) break; try { wait(); } catch( InterruptedException e ) { e.printStackTrace(); } } } // do the work ... // wait for everyone else to finish barrier.await(); } } } Again, this works well functionally (it does what it should). blocksize | system | user | cycles/sec 256k 1.9% 85% 1.30 64k 2.4% 80% 6.3 16k 4.6% 60% 30.1 4096 8.6% 41% 98.5 1024 12% 23% 202 256 14% 11.6% 299 64 14% 10.0% 518 16 14.8% 8.7% 679 The throughput for small work items is still much worse than that of the ExecutorService, but about 2x that of the CyclicBarrier. Eliminating one CyclicBarrier eliminates half of the gap. V4 - Busy wait instead of wait/notify Since this app is the primary one running on the system and the cores idle anyway if they're not busy with a work item, why not try a busy wait for work items in each thread, even if that spins the CPU needlessly. The worker thread code changes as follows: class WorkerThreadRunnable implements Runnable { // as before final protected void setWorkItem( @NotNull final Callable<Integer> callable ) { workItem = callable; } public void run() { while( true ) { // busy-wait for work while( true ) { if( workItem != NO_WORK ) break; } // do the work ... // wait for everyone else to finish barrier.await(); } } } Also works well functionally (it does what it should). blocksize | system | user | cycles/sec 256k 1.9% 85% 1.30 64k 2.2% 81% 6.3 16k 4.2% 62% 33 4096 7.5% 40% 107 1024 10.4% 23% 210 256 12.0% 12.0% 310 64 11.9% 10.2% 550 16 12.2% 8.6% 741 For small work items, this increases throughput by a further 10% over the CyclicBarrier + wait/notify variant, which is not insignificant. But it is still much lower-throughput than V1 with the ExecutorService. V5 - ? So what is the best synchronization mechanism for such a (presumably not uncommon) problem? I am weary of writing my own sync mechanism to completely replace ExecutorService (assuming that it is too generic and there has to be something that can still be taken out to make it more efficient). It is not my area of expertise and I'm concerned that I'd spend a lot of time debugging it (since I'm not even sure my wait/notify and busy wait variants are correct) for uncertain gain. Any advice would be greatly appreciated.

    Read the article

  • How Should I Generate Trade Statistics For CouchDB/Rails3 Application?

    - by James
    My Problem: I am trying to developing a web application for currency traders. The application allows traders to enter or upload information about their trades and I want to calculate a wide variety of statistics based on what the user entered. Now, normally I would use a relational database for this, but I have two requirements that don't fit well with a relational database so I am attempting to use couchdb. Those two problems are: 1) Primarily, I have a companion desktop application that users will be able to work with and replicate to the site using couchdb's awesome replication feature and 2) I would like to allow users to be able to define their own custom things to track about trades and generate results based off of what they enter. The schema less nature of couch seems perfect here, but it may end up being harder than it sounds. (I already know couch requires you to define views in advance and such so I was just planning on sticking all the custom attributes in an array and then emitting the array in the view and further processing from there.) What I Am Doing: Right now I am just emitting each trade in couch keyed by each user's system and querying with the key of the system to get an array of trades per system. Simple. I am not using a reduce function currently to calculate any stats because I couldn't figure out how to get everything I need without getting a reduce overflow error. Here is an example of rows that are getting emitted from couch: {"total_rows":134,"offset":0,"rows":[ {"id":"5b1dcd47221e160d8721feee4ccc64be", "key":["80e40ba2fa43589d57ec3f1d19db41e6","2010/05/14 04:32:37 +0000"], null, "doc":{ "_id":"5b1dcd47221e160d8721feee4ccc64be", "_rev":"1-bc9fe763e2637694df47d6f5efb58e5b", "couchrest-type":"Trade", "system":"80e40ba2fa43589d57ec3f1d19db41e6", "pair":"EUR/USD", "direction":"Buy", "entry":12600, "exit":12700, "stop_loss":12500, "profit_target":12700, "status":"Closed", "slug":"101332132375", "custom_tracking": [{"name":"signal", "value":"Pin Bar"}] "updated_at":"2010/05/14 04:32:37 +0000", "created_at":"2010/05/14 04:32:37 +0000", "result":100}} ]} In my rails 3 controller I am basically just populating an array of trades such as the one above and then extracting out the relevant data into smaller arrays that I can compute my statistics on. Here is my show action for the page that I want to display the stats and all the trades: def show @trades = Trade.by_system(:startkey => [@system.id], :endkey => [@system.id, Time.now ]) @trades.each do |trade| if trade.result > 0 @winning_trades << trade.result elsif trade.result < 0 @losing_trades << trade.result else @breakeven_trades << trade.result end if trade.direction == "Buy" @long_trades << trade.result else @short_trades << trade.result end if trade["custom_tracking"] != nil @custom_tracking << {"result" => trade.result, "variables" => trade["custom_tracking"]} end end end I am omitting some other stuff that is going on, but that is the gist of what I am doing. Then I am calculating stuff in the view layer to produce some results: <% winning_long_trades = @long_trades.reject {|trade| trade <= 0 } %> <% winning_short_trades = @short_trades.reject {|trade| trade <= 0 } %> <ul> <li>Total Trades: <%= @trades.count %></li> <li>Winners: <%= @winning_trades.size %></li> <li>Biggest Winner (Pips): <%= @winning_trades.max %></li> <li>Average Win(Pips): <%= @winning_trades.sum/@winning_trades.size %></li> <li>Losers: <%= @losing_trades.size %></li> <li>Biggest Loser (Pips): <%= @losing_trades.min %></li> <li>Average Loss(Pips): <%= @losing_trades.sum/@losing_trades.size %></li> <li>Breakeven Trades: <%= @breakeven_trades.size %></li> <li>Long Trades: <%= @long_trades.size %></li> <li>Winning Long Trades: <%= winning_long_trades.size %></li> <li>Short Trades: <%= @short_trades.size %></li> <li>Winning Short Trades: <%= winning_short_trades.size %></li> <li>Total Pips: <%= @winning_trades.sum + @losing_trades.sum %></li> <li>Win Rate (%): <%= @winning_trades.size/@trades.count.to_f * 100 %></li> </ul> This produces the following results, which aside from a few things is exactly what I want: Total Trades: 134 Winners: 70 Biggest Winner (Pips): 1488 Average Win(Pips): 440 Losers: 58 Biggest Loser (Pips): -516 Average Loss(Pips): -225 Breakeven Trades: 6 Long Trades: 125 Winning Long Trades: 67 Short Trades: 9 Winning Short Trades: 3 Total Pips: 17819 Win Rate (%): 52.23880597014925 What I Am Wondering- Finally The Actual Questions: I am starting to get really skeptical of how well this method will work when a user has 5,000 trades instead of just 134 like in this example. I anticipate most users will only have somewhere under 200 per year, but some users may have a couple thousand trades per year. Probably no more than 5,000 per year. It seems to work ok now, but the page load times are already getting a tad high for my tastes. (About 800ms to generate the page according to rails logs with about a 250ms of that spent in the view layer.) I will end up caching this page I am sure, but I still need the regenerate the page each time a trade is updated and I can't afford to have this be too slow. Sooo..... Is doing something similar here possible with a straight couchdb reduce function? I am assuming handing this off to couch would possibly help with larger data sets. I couldn't figure out how, but I suppose that doesn't mean it isn't possible. If possible, any hints will be helpful. Could I use a list function if a reduce was not available due to reduce constraints? Are couchdb list functions suitable for this type of calculations? Anyone have any idea of whether or not list functions perform well? Any hints what one would look like for the type of calculations I am trying to achieve? I thought about other options such as running the calculations at the time each trade was saved or nightly if I had to and saving the results to a statistics doc that I could then query so that all the processing was done ahead of time. I would like this to be the last resort because then I can't really filter out trades by time periods dynamically like I would really like to. (I want to have a slider that a user can slide to only show trades from that time period using the startkey and endkey in couchdb if I can.) If I should continue running the calculations inside the rails app at the time of the page view, what can I do to improve my current implementation. I am new to rails, couch and programming in general. I am sure that I could be doing something better here. Do I need to create an array for each stat or is there a better way to do that. I guess I just would really like some advice on how to tackle this problem. I want to keep the page generation time minimal since I anticipate these being some of the highest trafficked pages. My gut is that I will need to offload the statistics calculation to either couch or run the stats in advance of when they are called, but I am not sure. Lastly: Like I mentioned above, one of the primary reasons for using couch is to allow users to define their own things to track per trade. Getting the data into couch is no problem, but how would I be able to take the custom_tracking array and find how many winning trades for each named tracking attribute. If anyone can give me any hints to the possibility of doing this that would be great. Thanks a bunch. Would really appreciate any help. Willing to fork out some $$$ if someone wants to take on the problem for me. (Don't know if that is allowed on stack overflow or not.)

    Read the article

  • Table sorting & pagination with jQuery and Razor in ASP.NET MVC

    - by hajan
    Introduction jQuery enjoys living inside pages which are built on top of ASP.NET MVC Framework. The ASP.NET MVC is a place where things are organized very well and it is quite hard to make them dirty, especially because the pattern enforces you on purity (you can still make it dirty if you want so ;) ). We all know how easy is to build a HTML table with a header row, footer row and table rows showing some data. With ASP.NET MVC we can do this pretty easy, but, the result will be pure HTML table which only shows data, but does not includes sorting, pagination or some other advanced features that we were used to have in the ASP.NET WebForms GridView. Ok, there is the WebGrid MVC Helper, but what if we want to make something from pure table in our own clean style? In one of my recent projects, I’ve been using the jQuery tablesorter and tablesorter.pager plugins that go along. You don’t need to know jQuery to make this work… You need to know little CSS to create nice design for your table, but of course you can use mine from the demo… So, what you will see in this blog is how to attach this plugin to your pure html table and a div for pagination and make your table with advanced sorting and pagination features.   Demo Project Resources The resources I’m using for this demo project are shown in the following solution explorer window print screen: Content/images – folder that contains all the up/down arrow images, pagination buttons etc. You can freely replace them with your own, but keep the names the same if you don’t want to change anything in the CSS we will built later. Content/Site.css – The main css theme, where we will add the theme for our table too Controllers/HomeController.cs – The controller I’m using for this project Models/Person.cs – For this demo, I’m using Person.cs class Scripts – jquery-1.4.4.min.js, jquery.tablesorter.js, jquery.tablesorter.pager.js – required script to make the magic happens Views/Home/Index.cshtml – Index view (razor view engine) the other items are not important for the demo. ASP.NET MVC 1. Model In this demo I use only one Person class which defines Person entity with several properties. You can use your own model, maybe one which will access data from database or any other resource. Person.cs public class Person {     public string Name { get; set; }     public string Surname { get; set; }     public string Email { get; set; }     public int? Phone { get; set; }     public DateTime? DateAdded { get; set; }     public int? Age { get; set; }     public Person(string name, string surname, string email,         int? phone, DateTime? dateadded, int? age)     {         Name = name;         Surname = surname;         Email = email;         Phone = phone;         DateAdded = dateadded;         Age = age;     } } 2. View In our example, we have only one Index.chtml page where Razor View engine is used. Razor view engine is my favorite for ASP.NET MVC because it’s very intuitive, fluid and keeps your code clean. 3. Controller Since this is simple example with one page, we use one HomeController.cs where we have two methods, one of ActionResult type (Index) and another GetPeople() used to create and return list of people. HomeController.cs public class HomeController : Controller {     //     // GET: /Home/     public ActionResult Index()     {         ViewBag.People = GetPeople();         return View();     }     public List<Person> GetPeople()     {         List<Person> listPeople = new List<Person>();                  listPeople.Add(new Person("Hajan", "Selmani", "[email protected]", 070070070,DateTime.Now, 25));                     listPeople.Add(new Person("Straight", "Dean", "[email protected]", 123456789, DateTime.Now.AddDays(-5), 35));         listPeople.Add(new Person("Karsen", "Livia", "[email protected]", 46874651, DateTime.Now.AddDays(-2), 31));         listPeople.Add(new Person("Ringer", "Anne", "[email protected]", null, DateTime.Now, null));         listPeople.Add(new Person("O'Leary", "Michael", "[email protected]", 32424344, DateTime.Now, 44));         listPeople.Add(new Person("Gringlesby", "Anne", "[email protected]", null, DateTime.Now.AddDays(-9), 18));         listPeople.Add(new Person("Locksley", "Stearns", "[email protected]", 2135345, DateTime.Now, null));         listPeople.Add(new Person("DeFrance", "Michel", "[email protected]", 235325352, DateTime.Now.AddDays(-18), null));         listPeople.Add(new Person("White", "Johnson", null, null, DateTime.Now.AddDays(-22), 55));         listPeople.Add(new Person("Panteley", "Sylvia", null, 23233223, DateTime.Now.AddDays(-1), 32));         listPeople.Add(new Person("Blotchet-Halls", "Reginald", null, 323243423, DateTime.Now, 26));         listPeople.Add(new Person("Merr", "South", "[email protected]", 3232442, DateTime.Now.AddDays(-5), 85));         listPeople.Add(new Person("MacFeather", "Stearns", "[email protected]", null, DateTime.Now, null));         return listPeople;     } }   TABLE CSS/HTML DESIGN Now, lets start with the implementation. First of all, lets create the table structure and the main CSS. 1. HTML Structure @{     Layout = null;     } <!DOCTYPE html> <html> <head>     <title>ASP.NET & jQuery</title>     <!-- referencing styles, scripts and writing custom js scripts will go here --> </head> <body>     <div>         <table class="tablesorter">             <thead>                 <tr>                     <th> value </th>                 </tr>             </thead>             <tbody>                 <tr>                     <td>value</td>                 </tr>             </tbody>             <tfoot>                 <tr>                     <th> value </th>                 </tr>             </tfoot>         </table>         <div id="pager">                      </div>     </div> </body> </html> So, this is the main structure you need to create for each of your tables where you want to apply the functionality we will create. Of course the scripts are referenced once ;). As you see, our table has class tablesorter and also we have a div with id pager. In the next steps we will use both these to create the needed functionalities. The complete Index.cshtml coded to get the data from controller and display in the page is: <body>     <div>         <table class="tablesorter">             <thead>                 <tr>                     <th>Name</th>                     <th>Surname</th>                     <th>Email</th>                     <th>Phone</th>                     <th>Date Added</th>                 </tr>             </thead>             <tbody>                 @{                     foreach (var p in ViewBag.People)                     {                                 <tr>                         <td>@p.Name</td>                         <td>@p.Surname</td>                         <td>@p.Email</td>                         <td>@p.Phone</td>                         <td>@p.DateAdded</td>                     </tr>                     }                 }             </tbody>             <tfoot>                 <tr>                     <th>Name</th>                     <th>Surname</th>                     <th>Email</th>                     <th>Phone</th>                     <th>Date Added</th>                 </tr>             </tfoot>         </table>         <div id="pager" style="position: none;">             <form>             <img src="@Url.Content("~/Content/images/first.png")" class="first" />             <img src="@Url.Content("~/Content/images/prev.png")" class="prev" />             <input type="text" class="pagedisplay" />             <img src="@Url.Content("~/Content/images/next.png")" class="next" />             <img src="@Url.Content("~/Content/images/last.png")" class="last" />             <select class="pagesize">                 <option selected="selected" value="5">5</option>                 <option value="10">10</option>                 <option value="20">20</option>                 <option value="30">30</option>                 <option value="40">40</option>             </select>             </form>         </div>     </div> </body> So, mainly the structure is the same. I have added @Razor code to create table with data retrieved from the ViewBag.People which has been filled with data in the home controller. 2. CSS Design The CSS code I’ve created is: /* DEMO TABLE */ body {     font-size: 75%;     font-family: Verdana, Tahoma, Arial, "Helvetica Neue", Helvetica, Sans-Serif;     color: #232323;     background-color: #fff; } table { border-spacing:0; border:1px solid gray;} table.tablesorter thead tr .header {     background-image: url(images/bg.png);     background-repeat: no-repeat;     background-position: center right;     cursor: pointer; } table.tablesorter tbody td {     color: #3D3D3D;     padding: 4px;     background-color: #FFF;     vertical-align: top; } table.tablesorter tbody tr.odd td {     background-color:#F0F0F6; } table.tablesorter thead tr .headerSortUp {     background-image: url(images/asc.png); } table.tablesorter thead tr .headerSortDown {     background-image: url(images/desc.png); } table th { width:150px;            border:1px outset gray;            background-color:#3C78B5;            color:White;            cursor:pointer; } table thead th:hover { background-color:Yellow; color:Black;} table td { width:150px; border:1px solid gray;} PAGINATION AND SORTING Now, when everything is ready and we have the data, lets make pagination and sorting functionalities 1. jQuery Scripts referencing <link href="@Url.Content("~/Content/Site.css")" rel="stylesheet" type="text/css" /> <script src="@Url.Content("~/Scripts/jquery-1.4.4.min.js")" type="text/javascript"></script> <script src="@Url.Content("~/Scripts/jquery.tablesorter.js")" type="text/javascript"></script> <script src="@Url.Content("~/Scripts/jquery.tablesorter.pager.js")" type="text/javascript"></script> 2. jQuery Sorting and Pagination script   <script type="text/javascript">     $(function () {         $("table.tablesorter").tablesorter({ widthFixed: true, sortList: [[0, 0]] })         .tablesorterPager({ container: $("#pager"), size: $(".pagesize option:selected").val() });     }); </script> So, with only two lines of code, I’m using both tablesorter and tablesorterPager plugins, giving some options to both these. Options added: tablesorter - widthFixed: true – gives fixed width of the columns tablesorter - sortList[[0,0]] – An array of instructions for per-column sorting and direction in the format: [[columnIndex, sortDirection], ... ] where columnIndex is a zero-based index for your columns left-to-right and sortDirection is 0 for Ascending and 1 for Descending. A valid argument that sorts ascending first by column 1 and then column 2 looks like: [[0,0],[1,0]] (source: http://tablesorter.com/docs/) tablesorterPager – container: $(“#pager”) – tells the pager container, the div with id pager in our case. tablesorterPager – size: the default size of each page, where I get the default value selected, so if you put selected to any other of the options in your select list, you will have this number of rows as default per page for the table too. END RESULTS 1. Table once the page is loaded (default results per page is 5 and is automatically sorted by 1st column as sortList is specified) 2. Sorted by Phone Descending 3. Changed pagination to 10 items per page 4. Sorted by Phone and Name (use SHIFT to sort on multiple columns) 5. Sorted by Date Added 6. Page 3, 5 items per page   ADDITIONAL ENHANCEMENTS We can do additional enhancements to the table. We can make search for each column. I will cover this in one of my next blogs. Stay tuned. DEMO PROJECT You can download demo project source code from HERE.CONCLUSION Once you finish with the demo, run your page and open the source code. You will be amazed of the purity of your code.Working with pagination in client side can be very useful. One of the benefits is performance, but if you have thousands of rows in your tables, you will get opposite result when talking about performance. Hence, sometimes it is nice idea to make pagination on back-end. So, the compromise between both approaches would be best to combine both of them. I use at most up to 500 rows on client-side and once the user reach the last page, we can trigger ajax postback which can get the next 500 rows using server-side pagination of the same data. I would like to recommend the following blog post http://weblogs.asp.net/gunnarpeipman/archive/2010/09/14/returning-paged-results-from-repositories-using-pagedresult-lt-t-gt.aspx, which will help you understand how to return page results from repository. I hope this was helpful post for you. Wait for my next posts ;). Please do let me know your feedback. Best Regards, Hajan

    Read the article

< Previous Page | 56 57 58 59 60 61 62 63 64 65 66 67  | Next Page >