Search Results

Search found 20275 results on 811 pages for 'general performance'.

facebook twitter digg google delicious technorati stumbleupon myspace wordpress linkedin gmail igoogle windows live tumblr viadeo yahoo buzz yahoo mail yahoo bookmarks favorites email print

Page 80/811 | < Previous Page | 76 77 78 79 80 81 82 83 84 85 86 87  | Next Page >

  • Performance Overhead of Perf Event Subsystem in Linux Kernel

    - by Bo Xiao
    Performance counters for Linux are a new kernel-based subsystem that provide a framework for all things performance analysis. It covers hardware level (CPU/PMU, Performance Monitoring Unit) features and software features (software counters, tracepoints) as well. Since 2.6.33, the kernel provide 'perf_event_create_kernel_counter' kernel api for developers to create kernel counter to collect system runtime information. What I concern most is the performance impact on overall system when tracepoint/ftrace is enabled. There are no docs I can find about them. I was once told that ftrace was implemented by dynamically patching code, will it slow the system dramatically?

    Read the article

  • Retaining Managed objects - more general retaining objects

    - by Luuk D. Jansen
    A quick question regarding Managed Objects. I created an Array with Managed Objects (in Object 1: TableViewConbtroller), and pass one of those objects to another class/object (object 2: TableCell). The original array should still be retained in the original caller class. Then Object 2 is released, does that mean that that particular item in the array is released as well, as the reference to it in Object 2 was released? I am trying to better understand how to work with ManagedObjects as I get 'Object was released' errors. [EDIT] After some experimenting I came across the following scenario: I have the main AppDelegate. In a different class I create an AppDelegate to obtain the ManagedObjectContext. appDelegate = (iDomsAppDelegate *)[[UIApplication sharedApplication] delegate]; [self setContext:[appDelegate managedObjectContext]]; When the class is finished, and I release it, the variable in the class 'appDelegate' is also released. But then the ManagedObjectContext is closed, and obvious any future attempt to use it will cause a crash. So should I leave the appDelegate unreleased? This comes to the same question as the above about when and how to release in those situations where an objects is used from another class. I think a way of putting it is, how to know when you own an object and when not.

    Read the article

  • A general question about compilation and interpretation.

    - by wucnuc
    Hi stackoverflow, I apologize in advance for the possible stupidity of this question. However, the following has been the source of some confusion for me and I know the people here will be able to handily clear up the confusion for me. Basically, I would like to finally understand the relationship between any and all of the following terms. Some of the terms I do actually understand pretty well, but some of them are similar in my mind and I would like to once and for all to see their relationships/distinctions laid out all at once. They are: compiler interpreter bytecode machine code assembler assembly language binary object code executable Ideally, an answer would use examples from Java and C++ and other well-known programming languages that a young-ish student like me would be familiar with. Also, if you want to throw in any other useful terms that would be fine too :)

    Read the article

  • Cross-platform general purpose C++ RPC library

    - by iUm
    Here's the task: Imagine, we have an applications and a plug-in for it (dynamic library). Interface between the application and the plug-in is completely defined. Now I need to run the application and the plug-in on different computers. I wrote a stub for the plug-in on a computer where the real applications is running. And the application loads it and calls its functions like if it were a native plug-in. On the other computer there's a stub instead of the real application, which loads the native plug-in. Now I need to organize RPCs between my stubs over the network, regardless the very transport. Usually, it's not difficult. But there're some restrictions: Application-plug-in interaction can be reenterable (e.g. application calls f1() from plug-in, in f1() plug-in calls g1() from application, in g1() application calls f2() from plug-in and so on...) Any such reenteration should be executed exactly by the same thread, which started the sequence Where can I find a cross-platform C++ RPC library with such features?

    Read the article

  • Me As Child Type In General Function

    - by Steven
    I have a MustInherit Parent class with two Child classes which Inherit from the Parent. How can I use (or Cast) Me in a Parent function as the the child type of that instance? EDIT: My actual goal is to be able to serialize (BinaryFormatter.Serialize(Stream, Object)) either of my child classes. However, "repeating the code" in each child "seems" wrong. EDIT2: This is my Serialize function. Where should I implement this function? Copying and pasting to each child doesn't seem right, but casting the parent to a child doesn't seem right either. Public Function Serialize() As Byte() Dim bFmt As New BinaryFormatter() Dim mStr As New MemoryStream() bFmt.Serialize(mStr, Me) Return mStr.ToArray() End Function

    Read the article

  • General workflow to allow multiple OpenIDs to be associated with one app account

    - by BobTodd
    I have a (typical?) scenario: that my app's users can use multiple openids mapped to one app account (like stackoverflow). For me the unique thing on the account is the email address, so this binds openids to the profile. Question is, how to allow a user to start using a second openid once one is setup. I am asking as I have read that it is a security hole to allow automatic account openid syncing simply based on the provider-supplied email address as someone could easily spoof someone's email address to create a spoof openid and falsely access the account (how I am not sure) - although this seems to be exactly how stack operates. See options a. and b. below. Problem for me with a. is what happens if the original openid no longer works for whatever reason - how would you set-up a new openid? Would b. be more acceptable if we used email verification? Does anyone have an article detailing a "standard" way (set of user stories) for this - it seems to be an increasingly popular way to authenticate. I have tried to detail this in a rough decision tree... 1. My Site > authentication landing page - user chooses an openid (facebook, google, myopenid etc), redirection > 2. Provider site returns with token (includes user registering a new openid, logging in or is already logged in to Provider site) 3. My Site > use token id to lookup user 3.1 Profile exists? Yes > authenticate. ends. No > 3.1.1 was email address supplied by provider? Yes > lookup user by email address 3.1.1.1 Profile exists? Yes > a. error message - please login with existing openid and associate this openid (from special page) Yes > b. or associate this openid with existing profile automatically. authenticate. ends. No > Register profile. With registration email address follow 3.1.1, except this time where email is unique, we will associate openid. ends

    Read the article

  • What are the general disadvantages to most frameworks?

    - by JamieC
    I'm trying to come up with a list of negative reasons for the use of frameworks. I know there are many advantages, such as partially complete solutions (Collections, Exception handling, etc). What disadvantages are there apart from the inverse of the above (ie: 'you will have to create your own collections/ own exception handling'). Thanks

    Read the article

  • Is there a general-purpose printf-ish routine defined in any C standard

    - by supercat
    In many C libraries, there is a printf-style routine which is something like the following: int __vgprintf(void *info, (void)(*print_function(void*, char)), const char *format, va_list params); which will format the supplied string and call print_function with the passed-in info value and each character in sequence. A function like fprintf will pass __vgprintf the passed-in file parameter and a pointer to a function which will cast its void* to a FILE* and output the passed-in character to that file. A function like snprintf will create a struct holding a char* and length, and pass the address of that struct to a function which will output each character in sequence, space permitting. Is there any standard for such a function, which could be used if e.g. one wanted a function to output an arbitrary format to a TCP port? A common approach is to allocate a buffer one hopes is big enough, use snprintf to put the data there, and then output the data from the buffer. It would seem cleaner, though, if there were a standard way to to specify that the print formatter should call a user-supplied routine with each character.

    Read the article

  • General Question about MD5

    - by chicane
    So im just playing around with PHP and the MD5 functionality, sorry if this sounds really silly, but I cant seem to understand, how is it possible to represent an unlimited number of characters of input into a 32 bit character output? Is my logic sound here? Or is there a limit to the input that a MD5 function can take? Thanks...

    Read the article

  • General advice from people in the industry - new graduate

    - by confusified
    I'm 20 years old and have just finished a 4 year Information Technology degree in Ireland, The main focus of the course was programming (mainly java) and software engineering. My question (posted in the wrong place as it may be) is : What technologies that I may not have studied should I attempt to teach myself that will be of the most benefit to me in searching for employment? All input appreciated.

    Read the article

  • Java 1.4 Class performance on 1.5 JVM

    - by user222164
    Switching from JVM 1.4 to 1.5 has performance benefits as per release notes. http://java.sun.com/j2se/1.5.0/docs/relnotes/features.html#performance We have Java 1.4 compiled classes which are run on 1.5 JVM, will these classes suffer in performance because they were compiled using 1.4 ?

    Read the article

  • Building Admin Areas in Rails - General Questions

    - by Carb
    What is the typical format/structure for creating an administrative area in a Rails application? Specifically I am stumped in the vicinity of these topics: How do you deal with situations where a model's resources are available to both the public and the Admin? i.e. A User model where anyone can create users, login, etc but only the admin can view users, delete/update them, etc. What is the proper convention for routing? How does one structure controllers? Are duplicate controllers considered OK? i.e. An admin version and the non-admin version? Thank you!

    Read the article

  • Linq-to-SQL and Performance.

    - by jalpesh
    HI, I am developing asp.net mvc site with linq-to-sql we are having 1000 cocurrent users and we are having performance problems. I have found that stackovewflow is also build on linq-to-sql? So can anybody know how they improved performance. Without line performance was good each page are loaded in 3 seconds but after migrating to linq as per our client requirement page comes in 8 to 10 seconds which is not acceptable performance. Our HTML is very clear but we are having very complex database operations. Any tip or code will be best answer. Thanks in advance,

    Read the article

  • General question about Ruby Eigenclass/Singleton

    - by Dex
    module MyModule def my_method; 'hello'; end end class MyClass class << self include MyModule end end MyClass.my_method # => "hello I'm unsure why "include MyModule" needs to be in the eigenclass in order to be called using just MyClass. Why can't I go: X = MyClass.new X.my_method How is this type of pattern useful if I can't call it from an instance?

    Read the article

  • General web service ideas

    - by user2014175
    I have a question regarding different types of web services. I'll preface this by saying that I have built a number of apps (for both ios and android) for personal use that interact with the web via php and sql. I have taught myself these languages, and as such don't have the broader background knowledge that many of you do. My question is, in what other ways can you perform an interaction between a web service and a mobile device other than mobile - php - sql - etc. For example, If I built a very simple tracking app for my car, my current method would be to push GPS coordinates from my iphone to my database at a set interval, then I would write a simple bit of javascript that pulled those coordinates out of the database and superimposed them on a google map. Is there a different way to do this? Such as the server acting as a live middle man who simple pushed the coordinates directly to a target browser? Without the database in the middle? If so, are there advantages and disadvantages to these different methods to achieve different goals? I know its a broad question but I'm really intrigued and I'm finding it difficult to word a google search for it. Any info / reading material suggesting would be excellent. Thanks

    Read the article

  • File sorting, Comparison method violates its general contract

    - by user2677383
    My file sorting comparator sometimes invoke java.lang.IllegalArgumentException in android app. I cannot figure out the reason. could you explain that? here is my code block: Comparator cc = new Comparator<File>() { @Override public int compare(File f1, File f2) { if (f1.isDirectory() && !f2.isDirectory()) { return 1; } if (f2.isDirectory() && !f1.isDirectory()) { return -1; } if ((f1.isFile() && f2.isFile()) || (f1.isDirectory() && f2.isDirectory())) { if (sort == 1) { return Util.compareFileName(f2.getName(), f1.getName()); } else if (sort == 2) { return (int) (f1.lastModified() - f2.lastModified()); } else if (sort == 3) { return (int) (f2.lastModified() - f1.lastModified()); } else if (sort == 4) { return (int) (f1.length() - f2.length()); } else if (sort == 5) { return (int) (f2.length() - f1.length()); } else return Util.compareFileName(f1.getName(), f2.getName()); } return f1.compareTo(f2); } }; Util.compareFileName is as followings: public static int compareFileName(String s1, String s2) { int thisMarker = 0; int thatMarker = 0; int s1Length = s1.length(); int s2Length = s2.length(); while (thisMarker < s1Length && thatMarker < s2Length) { String thisChunk = getChunk(s1, s1Length, thisMarker); thisMarker += thisChunk.length(); String thatChunk = getChunk(s2, s2Length, thatMarker); thatMarker += thatChunk.length(); // If both chunks contain numeric characters, sort them numerically int result = 0; if (isDigit(thisChunk.charAt(0)) && isDigit(thatChunk.charAt(0))) { // Simple chunk comparison by length. int thisChunkLength = thisChunk.length(); result = thisChunkLength - thatChunk.length(); // If equal, the first different number counts if (result == 0) { for (int i = 0; i < thisChunkLength; i++) { result = thisChunk.charAt(i) - thatChunk.charAt(i); if (result != 0) { return result; } } } } else { result = thisChunk.compareTo(thatChunk); } if (result != 0) return result; } return s1Length - s2Length; }

    Read the article

  • fastest way to perform string search in general and in python

    - by Rkz
    My task is to search for a string or a pattern in a list of documents that are very short (say 200 characters long). However, say there are 1 million documents of such time. What is the most efficient way to perform this search?. I was thinking of tokenizing each document and putting the words in hashtable with words as key and document number as value, there by creating a bag of words. Then perform the word search and retrieve the list of documents that contained this word. From what I can see is this operation will take O(n) operations. Is there any other way? may be without using hash-tables?. Also, is there a python library or third party package that can perform efficient searches?

    Read the article

  • Netbook performance - 1.33 GHz vs 1.6/1.66 GHz Atom

    - by Imran
    All new 11" netbooks seem to carry 1.33 GHz Atom Z520 CPU instead of 1.6/1.66 GHz Atom N270/N280. The screen resolution of 11" netbooks make them very appealing, but I'm a bit concerned about their performance as they carry a slower CPU than the 1.6GHz Atom, which isn't a great performer in the first place. Is there any significant difference in performance between 1.33 GHz and 1.6/1.66 GHz Atom processors in day to day usage? Are any of those fast enough to decode 720p x264 video? (When paired with typical Intel GMA platform and software decoder like ffdshow/CoreAVC of course, not with Nvidia Ion platform)

    Read the article

  • Any reason not to disable Windows kernel paging?

    - by Nathaniel
    So I'm planning on eventually going to 2 GB (mobo max) RAM from 1 GB, and I want to disable kernel paging once I do, because I've heard it can give a performance boost (and that I believe). Any reason not to do it or any general thoughts about it? Edit: for clarification, this is not disabling general RAM paging. This is disabling having kernel memory paged (or at least parts of it, as Charlls noted).

    Read the article

  • MD3200 - 3 to 4x less throughput than MD1220. Am I missing something here?

    - by Igor Polishchuk
    I have two R710 servers with similar configuration. One in my office has MD1220 attached. Another one in the datacenter of my hosting services vendor has MD3200. I'm getting significantly worse throughput from MD3200 at my vendors setup. I'm mostly interested in sequential writes, and I'm getting these results in bonnie++ and dd tests: Seq. writes on MD1220 in my office: 1.1 GB/s - bonnie++, 1.3GB/s - dd Seq. writes on MD3200 at my vendor's: 240MB/s - bonnie++, 310MB/s - dd Unfortunately, I could not test the exactly the same configurations, but the two I have should be comparable. If anything, my good performing environment is cheaper than the bad performing. I expect at least similar throughput from these two setups. My vendor cannot really help me. Hopefully, somebody more familiar with the DAS performance can look at it and tell if I'm missing something here and my expectations are too high. To summarize, the question here is it reasonable to expect about 100MB/s of sequential write throughput per each couple of drives in RAID10 on MD3200? Is there any trick to enable such performance in MD3200 with dual controller as opposed to simple MD1220 with a single H800 adapter? More details about the configurations: A good one in my office: Dell R710 2CPU X5650 @ 2.67GHz 12 cores 96GB DDR3, OS: RHEL 5.5, kernel 2.6.18-194.26.1.el5 x86_64 20x300GB 2.5" SAS 10K in a single RAID10 1MB chunk size on MD1220 + Dell H800 I/O controller with 1GB cache in the host Not so good one at my vendor's: Dell R710 2CPU L5520 @ 2.27GHz 8 cores 144GB DDR3, OS: RHEL 5.5, kernel 2.6.18-194.11.4.el5 x86_64 20x146GB 2.5" SAS 15K in a single RAID10 512KB chunk size, Dell MD3200, 2 I/O controllers in array with 1GB cache each Additional information. I've also ran the same tests on the same vendor's host, but the storage was: two raids of 14x146GB 15K RPM drives RAID 10, striped together on the OS level on MD3000+MD1000. The performance was about 25% worse than on MD3200 despite having more drives. When I ran similar tests on the internal storage of my vendor's host (2x146GB 15K RPM drives RAID1, Perc 6i) I've got about 128MB/s seq. writes. Just two internal drives gave me about a half of 20 drives' throughput on MD3200. The random I/O performance of the MD3200 setup is ok, it gives me at least 1300 IOPS. I'm mostly have problems with sequentioal I/O throughput. Thank you for looking into it. Regards Igor

    Read the article

facebook twitter digg google delicious technorati stumbleupon myspace wordpress linkedin gmail igoogle windows live tumblr viadeo yahoo buzz yahoo mail yahoo bookmarks favorites email print

< Previous Page | 76 77 78 79 80 81 82 83 84 85 86 87  | Next Page >