Search Results

Search found 13889 results on 556 pages for 'results'.

Page 72/556 | < Previous Page | 68 69 70 71 72 73 74 75 76 77 78 79  | Next Page >

  • How do I find information on who links to my sites?

    - by bobdobbs
    I'm trying to figure out if there's a free way to get information on backlinks to my site. I've had webmaster tools and google analytics set up for years. But I can't find access to data about site backlinks in either toolset. Webmaster tools, under 'traffic'-'links to your site' gives me the same message for all of my sites: "No data available". I haven't been able to find anything in GA that gives any information on backlinks. I've heard of using "links:" as an operator in google search, but for each of my sites, this returns either zero or very few results in cases when I know I have many backlinks. Most of the links simple aren't shown. My thinking is that google maintains a graph of who links to my site, so I figured that they might let me see it. But I can't figure out how. I've found this tool on a spammy website: http://www.backlinkwatch.com. It offers more data than google on my backlines, and offers more results in exchange for a paid subscription. The data it offers for free looks good, but the results are limited and the site has popups and obnoxious ads. So, in short: how do I get data on who links to me? Is there a free way?

    Read the article

  • SharePoint: UI for ordering list items

    - by svdoever
    SharePoint list items have in the the base Item template a field named Order. This field is not shown by default. SharePoint 2007, 2010 and 2013 have a possibility to specify the order in a UI, using the _layouts page: {SiteUrl}/_layouts/Reorder.aspx?List={ListId} In SharePoint 2010 and 2013 it is possible to add a custom action to a list. It is possible to add a custom action to order list items as follows (SharePoint 2010 description): Open SharePoint Designer 2010 Navigate to a list Select Custom Actions > List Item Menu Fill in the dialog box: Open List Settings > Advanced Settings > Content Types, and set Allow management of content types to No  On List Settings select Column Ordering This results in the following UI in the browser: Selecting the custom Order Items action (under List Tools > Items) results in: You can change your custom action in SharePoint designer. On the list screen in the bottom right corner you can find the custom action: We now need to modify the view to include the order by adding the Order field to the view, and sorting on the Order field. Problem is that the Order field is hidden. It is possible to modify the schema of the Order field to set the Hidden attribute to FALSE. If we don’t want to write code to do it and still get the required result it is also possible to modify the view through SharePoint Designer: Modify the code of the view: This results in: Note that if you change the view through the web UI these changes are partly lost. If this is a problem modify the Order field schema for the list.

    Read the article

  • How do I throttle a command in a terminal window?

    - by To Do
    I needed to run convert with a lot of images at the same time. The command took quite a while but this doesn't bother me. The issue is that this command rendered my computer unusable while the command was running (for about 15 minutes). So is it possible to throttle the command by limiting resources (processor and memory) to the command, directly from the command line? This can only work if I add something to the same line before pressing Enter because once I start the process the computer slows so much that it is impossible for example to switch to "System monitor" and reduce priority. Edit: top and iotop results I managed to run top and sudo iotop >iotop.txt while doing one of these convert operations. (The iotop.txt file produced is difficult to read) Results of top: PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 14275 username 20 0 4043m 3.0g 1448 D 7.0 80.4 0:16.45 convert Results of iotop: [?1049h[1;24r(B[m[4l[?7h[?1h=[39;49m[?25l[39;49m(B[m[H[2JTotal DISK READ: 1269.04 K/s | Total DISK WRITE:[59G0.00 B/s (B[0;7m TID PRIO USER DISK READ DISK WRITE SWAPIN(B[0;1;7m IO(B[0;7m COMMAND [3;2H(B[m2516 be/4 username 350.08 K/s 0.00 B/s 0.00 % 0.00 % zeitgeist-datahub 7394 be/4 username 568.88 K/s 0.00 B/s 77.41 % 0.00 % --rendere~.530483991[5;1H14275 idle username 350.08 K/s 0.00 B/s 37.49 % 0.00 % convert S~f test.pdf[6;2H2048 be/4 root[6;24H0.00 B/s 0.00 B/s 0.00 % 0.00 % [kworker/3:2] [5G1 be/4 root[7;24H0.00 B/s 0.00 B/s 0.00 % 0.00 % init Furthermore, even after the process ends, the computer does not return to the previous performance. I found a way around this by running sudo swapoff -a followed by sudo swapon -a

    Read the article

  • How important is the uniqueness of your domain name?

    - by Corey
    I've finally come up with a domain name that I like and is available. The name is nonsensical and doesn't translate into anything meaningful in any language, as far as I know. It's something like "FOOBARite". (Don't steal that!) I'm wondering about a few search issues. Results-wise, searching for it in Google currently returns about 15k results, none of which are relevant (dead Twitter pages, various unpopular online handles, and botched french translations). However, Google starts off with a spelling suggestion, which removes a letter. ("Did you mean: FOOBARit?") That returns about 250k results for several different and unrelated websites/organizations by that name. One is some technology provider, another is a sign-language organization, another is the name of a font... None of them seem particularly popular, there's not that much activity on any of those pages. Anyway, the two are pronounced differently, they're just a letter off. Should I go with my idea or is this one-letter variation going to cause me problems? If my site becomes ranked well enough, will Google's spelling suggestion go away? I don't want users to search for my site name and be told they've spelled it wrong.

    Read the article

  • Force Windows 8 to search indexed files

    - by Hrvoje
    When using search files in Windows 8 (win+f) I don't get expected results. For example, I installed VLC, it's in Program Files (86) folder, and that folder is selected for indexing. Search for files (win+f) gives 0 results. If I pin to start that exe, then it's found - but I don't want to do this, that's not the point. Where does it search for files? Is there any way to specify search locations? It doesn't use Indexing Options settings, at least it seams so. Also, searching from explorer window is kinda slow - I tried entering VLC.EXE in search box (when in c:\ root), and it takes some time to give correct results. It works, but it looks like it doesn't use indexing, rather scan all files/folders, which is slow.

    Read the article

  • Can I use a 302 redirect to serve up static content from an URL with escaped_fragment?

    - by Starfs
    We would like to serve up SEO-friendly Ajax-driven content. We are following this documentation. Has anyone ever tried to write a 302 redirect into the .htaccess file, that takes the ?_escaped_fragment= string and send that to a static page?, for example /snapshot/yourfilename/. How will Google react to this? I've gone through the documentation and it's not very clear. The below quote is from Google's documentation this is what I find. I'm not sure if they are saying that you can redirect the _escaped_fragment_ URL to a different static page, or if this is to redirect the hashtag URL to static content? Thoughts? From Google's site: Question: Can I use redirects to point the crawler at my static content? Redirects are okay to use, as long as they eventually get you to a page that's equivalent to what the user would see on the #! version of the page. This may be more convenient for some webmasters than serving up the content directly. If you choose this approach, please keep the following in mind: Compared to serving the content directly, using redirects will result in extra traffic because the crawler has to follow redirects to get the content. This will result in a somewhat higher number of fetches/second in crawl activity. Note that if you use a permanent (301) redirect, the url shown in our search results will typically be the target of the redirect, whereas if a temporary (302) redirect is used, we'll typically show the #! url in search results. Depending on how your site is set up, showing #! may produce a better user experience, because the user will be taken straight into the AJAX experience from the Google search results page. Clicking on a static page will take them to the static content, and they may experience avoidable extra page load time if the site later wants to switch them to the AJAX experience.

    Read the article

  • How do I interpolate air drag with a variable time step?

    - by Valentin Krummenacher
    So I have a little game which works with small steps, however those steps vary in time, so for example I sometimes have 10 Steps/second and then I have 20 Steps/second. This changes automatically depending on how many steps the user's computer can take. To avoid inaccurate positioning of the game's player object I use y=v0*dt+g*dt^2/2 to determine my objects y-position, where dt is the time since the last step, v0 is the velocity of my object in the beginning of my step and g is the gravity. To calculate the velocity in the end of a step I use v=v0+g*dt what also gives me correct results, independent of whether I use 2 steps with a dt of for example 20ms or one step with a dt of 40ms. Now I would like to introduce air drag. For simplicity's sake I use a=k*v^2 where a is the air drag's acceleration (I am aware that it would usually result in a force, but since I assume 1kg for my object's mass the force is the same as the resulting acceleration), k is a constant (in this case I'm using 0.001) and v is the speed. Now in an infinitely small time interval a is k multiplied by the velocity in this small time interval powered by 2. The problem is that v in the next time interval would depend on the drag of the last which again depends on the v of the last interval and so on... In other words: If I use a=k*v^2 I get different results for my position/velocity when I use 2 steps of 20ms than when I use one step of 40ms. I used to have this problem for my position too, but adding +g*dt^2/2 to the formula for my position fixed the problem since it takes into account that the position depends on the velocity which changes slightly in every infinitely small time interval. Does something like that exist for air drag too? And no, I dont mean anything like Adding air drag to a golf ball trajectory equation or similar, for that kind of method only gives correct results when all my steps are the same. (I hope you can understand my intermediate english, it's not my main language so I would like to say sorry for all the silly mistakes I might have made in my question)

    Read the article

  • Using Bulk Operations with Coherence Off-Heap Storage

    - by jpurdy
    Some NamedCache methods (including clear(), entrySet(Filter), aggregate(Filter, …), invoke(Filter, …)) may generate large intermediate results. The size of these intermediate results may result in out-of-memory exceptions on cache servers, and in some cases on cache clients. This may be particularly problematic if out-of-memory exceptions occur on more than one server (since these operations may be cluster-wide) or if these exceptions cause additional memory use on the surviving servers as they take over partitions from the failed servers. This may be particularly problematic with clusters that use off-heap storage (such as NIO or Elastic Data storage options), since these storage options allow greater than normal cache sizes but do nothing to address the size of intermediate results or final result sets. One workaround is to use a PartitionedFilter, which allows the application to break up a larger operation into a number of smaller operations, each targeting either a set of partitions (useful for reducing the load on each cache server) or a set of members (useful for managing client result set sizes). It is also possible to return a key set, and then pull in the full entries using that key set. This also allows the application to take advantage of near caching, though this may be of limited value if the result is large enough to result in near cache thrashing.

    Read the article

  • Email Discovery from Fairly Large Mailbox (15gig) Exchange 2003.

    - by nysingh
    I have a request from our legal team to search a users' mailbox. the mailbox is 15gig and it is on exchange 2003. I am trying to run windows desktop search and google desktop. I have gotten them to index mailbox but getting the results into a folder to backup on cd is getting bit difficult. Windows desktop search and google desktop search does not allow you to copy results to another folder. Can anyone point me to right direction? What is the best way to index and copy the results of pst, mailbox or edb file? What is the best discovery methods? Thanks

    Read the article

  • How to search the web for programming related solutions?

    - by Bob
    I have the impression that Google has become unusable when searching for programming related questions. Example: I'm Googling for XML-RPC Redstone Cookie I'm expecting results where all three terms are contained. I don't care for results where one term misses. I guess until some months ago Google just worked this way, i.e. all terms were included. Somehow this feature is gone now (Google apparently thinks it is more intelligent than the user and knows what the user is searching for). So I helped myself putting a + in front of every word. This is, however, a bit cumbersome. And for the last weeks, it even doesn't work anymore in all cases, Google ignores the +. So how do you search for progamming related problems? Do you still use Google? If yes, which techniques do you use to get the right results? Or do you use another search engine? Which one?

    Read the article

  • Python Mechanize unable to avoid redirect when Post

    - by Enric Geijo
    I am trying to crawl a site using mechanize. The site provides search results in different pages. When posting to get the next set of results, something is wrong and the server redirects me to the first page, asking mechanize to update the SearchSession Cookie. I have been debugging the requests using Firefox and they look quite the same, and I am unable to find the problem. Any suggestion? Below the requests: ----------- FIRST THE RIGHT SEQUENCE, USING TAMPER IN FIREFOX ------------------------- POST XXX/JobSearch/Results.aspx?Keywords=Python&LTxt=London%2c+South+East&Radius=0&LIds2=ZV&clid=1621&cltypeid=2&clName=London Load Flags[LOAD_DOCUMENT_URI LOAD_INITIAL_DOCUMENT_URI ] Content Size[-1] Mime Type[text/html] Request Headers: Host[www.cwjobs.co.uk] User-Agent[Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.9) Gecko/20100401 Ubuntu/9.10 (karmic) Firefox/3.5.9] Accept[text/html,application/xhtml+xml,application/xml;q=0.9,/;q=0.8] Accept-Language[en-us,en;q=0.5] Accept-Encoding[gzip,deflate] Accept-Charset[ISO-8859-1,utf-8;q=0.7,*;q=0.7] Keep-Alive[300] Connection[keep-alive] Referer[XXX/JobSearch/Results.aspx?Keywords=Python&LTxt=London%2c+South+East&Radius=0&LIds2=ZV&clid=1621&cltypeid=2&clName=London] Cookie[ecos=774803468-0; AnonymousUser=MemberId=acc079dd-66b6-4081-9b07-60d6955ee8bf&IsAnonymous=True; PJBIPPOPUP=; WT_FPC=id=86.181.183.106-2262469600.30073025:lv=1272812851736:ss=1272812789362; SearchSession=SessionGuid=71de63de-3bd0-4787-895d-b6b9e7c93801&LogSource=NAT] Post Data: __EVENTTARGET[srpPager%24btnForward] __EVENTARGUMENT[] hdnSearchResults[BV%2CA%2CC0P5x%2COou-%2CB4S-%2CBuC-%2CDzx-%2CHwn-%2CKPP-%2CIVA-%2CC9D-%2CH6X-%2CH7x-%2CJ0x-%2CCvX-%2CCra-%2COHa-%2CHhP-%2CCoj-%2CBlM-%2CE9W-%2CIm8-%2CBqG-%2CPFy-%2CN%2Fm-%2Ceaa%2CCvj-%2CCtJ-%2CCr7-%2CBpu-%2Cmh%2CMb6-%2CJ%2Fk-%2CHY8-%2COJ7-%2CNtF-%2CEya-%2CErT-%2CEo4-%2CEKU-%2CDnL-%2CC5M-%2CCyB-%2CBsD-%2CBrc-%2CBpU-%2Col%2C30%2CC1%2Cd4N%2COo8-%2COi0-%2CLz%2F-%2CLxP-%2CFyp-%2CFVR-%2CEHL-%2CPrP-%2CLmE-%2CK3H-%2CKXJ-%2CFyn%2CIcq-%2CIco-%2CIK4-%2CIIg-%2CH2k-%2CH0N-%2CHwp-%2CHvF-%2CFij-%2CFhl-%2CCwj-%2CCb5-%2CCQj-%2CCQh-%2CB%2B2-%2CBc6-%2ChFo%2CNLq-%2CNI%2F-%2CFzM-%2Cdu-%2CHg2-%2CBug-%2CBse-%2CB9Q-] __VIEWSTATE[%2FwEPDwUKLTkyMzI2ODA4Ng9kFgYCBA8WBB4EaHJlZgWJAWh0dHA6Ly93d3cuY3dqb2JzLmNvLnVrL0pvYlNlYXJjaC9SU1MuYXNweD9LZXl3b3Jkcz1QeXRob24mTFR4dD1Mb25kb24lMmMrU291dGgrRWFzdCZSYWRpdXM9MCZMSWRzMj1aViZjbGlkPTE2MjEmY2x0eXBlaWQ9MiZjbE5hbWU9TG9uZG9uHgV0aXRsZQUkTGF0ZXN0IFB5dGhvbiBqb2JzIGZyb20gQ1dKb2JzLmNvLnVrZAIGDxYCHgRUZXh0BV48bGluayByZWw9ImNhbm9uaWNhbCIgaHJlZj0iaHR0cDovL3d3dy5jd2pvYnMuY28udWsvSm9iU2Vla2luZy9QeXRob25fTG9uZG9uX2wxNjIxX3QyLmh0bWwiIC8%2BZAIIEGRkFg4CBw8WAh8CBV9Zb3VyIHNlYXJjaCBvbiA8Yj5LZXl3b3JkczogUHl0aG9uOyBMb2NhdGlvbjogTG9uZG9uLCBTb3V0aCBFYXN0OyA8L2I%2BIHJldHVybmVkIDxiPjg1PC9iPiBqb2JzLmQCCQ8WAh4HVmlzaWJsZWhkAgsPFgIfAgUoVGhlIG1vc3QgcmVsZXZhbnQgam9icyBhcmUgbGlzdGVkIGZpcnN0LmQCEw8PFgIeC05hdmlnYXRlVXJsBQF%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%3D] Refinesearch%24txtKeywords[Python] Refinesearch%24txtLocation[London%2C+South+East] Refinesearch%24ddlRadius[0] ddlCompanyType[0] ddlSort[1] Response Headers: Cache-Control[private] Date[Sun, 02 May 2010 16:09:27 GMT] Content-Type[text/html; charset=utf-8] Expires[Sat, 02 May 2009 16:09:27 GMT] Server[Microsoft-IIS/6.0] X-SiteConHost[P310] X-Powered-By[ASP.NET] X-AspNet-Version[2.0.50727] Set-Cookie[SearchSession=SessionGuid=71de63de-3bd0-4787-895d-b6b9e7c93801&LogSource=NAT; path=/] Content-Encoding[gzip] Vary[Accept-Encoding] Transfer-Encoding[chunked] -------- NOW WHAT I'AM SENDING USING MECHANIZE, SOME HEADERS ADDED, ETC ----------- POST /JobSearch/Results.aspx?Keywords=Python&LTxt=London%2c+South+East&Radius=0&LIds2=ZV&clid=1621&cltypeid=2&clName=London HTTP/1.1\r\nContent-Length: 2424\r\n Accept-Language: en-us,en;q=0.5\r\n Accept-Encoding: gzip\r\n Host: www.cwjobs.co.uk\r\n Accept: text/html,application/xhtml+xml,application/xml;q=0.9,/;q=0.8\r\n Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n Connection: keep-alive\r\n Cookie: AnonymousUser=MemberId=8fa5ddd7-17ed-425e-b189-82693bfbaa0c&IsAnonymous=True; SearchSession=SessionGuid=33e4e439-c2d6-423f-900f-574099310d5a&LogSource=NAT\r\n Referer: XXX/JobSearch/Results.aspx?Keywords=Python&LTxt=London%2c+South+East&Radius=0&LIds2=ZV&clid=1621&cltypeid=2&clName=London\r\n Content-Type: application/x-www-form-urlencoded\r\n\r\n' '__EVENTTARGET=srpPager%24btnForward& __EVENTARGUMENT=& hdnSearchResults=BV%2CA%2CC0eif%2CMwc%2CM6s%2COou%2CK09%2CG4H%2CEZf%2CGTu%2CLrr%2CGuX%2CGs9%2CEz9%2CL5X%2CL9U%2ChU%2CHHf%2CMAL%2CNDi%2CJrY%2CGBy%2CM%2Bo%2CdE-%2CpI%2CtDI%2CL5L%2CL7l%2CL8z%2CM%2FA%2CPPP%2CCM0%2CEpK%2CHPy%2Cez%2C7p%2CJ2U%2CJ9b%2CJ%2F2%2CKea%2CLBj%2CLvi%2CL2t%2CM8r%2CM9S%2CM%2Fa%2CPRT%2CPgi%2Csg7%2CF6%2CI2F%2CJTd%2CO-%2CC0v%2CC3f%2CDCq%2CDxn%2CERl%2CUbV%2CGME%2CGMG%2CGd2%2CGgO%2CGyK%2CG0h%2CG4F%2CG5p%2CJGL%2CJHJ%2CKhj%2CL4L%2CMM1%2CMYL%2CMYN%2CMp4%2CNL0%2COrj%2CvuW%2CBdE%2CBfv%2CI1i%2CBCh-%2COLA%2CHH4%2CM6O%2CM8Q%2CMre& __VIEWSTATE=%2FwEPDwUKLTkyMzI2ODA4Ng9kFgYCBA8WBB4EaHJlZgWJAWh0dHA6Ly93d3cuY3dqb2JzLmNvLnVrL0pvYlNlYXJjaC9SU1MuYXNweD9LZXl3b3Jkcz1QeXRob24mTFR4dD1Mb25kb24lMmMrU291dGgrRWFzdCZSYWRpdXM9MCZMSWRzMj1aViZjbGlkPTE2MjEmY2x0eXBlaWQ9MiZjbE5hbWU9TG9uZG9uHgV0aXRsZQUkTGF0ZXN0IFB5dGhvbiBqb2JzIGZyb20gQ1dKb2JzLmNvLnVrZAIGDxYCHgRUZXh0BV48bGluayByZWw9ImNhbm9uaWNhbCIgaHJlZj0iaHR0cDovL3d3dy5jd2pvYnMuY28udWsvSm9iU2Vla2luZy9QeXRob25fTG9uZG9uX2wxNjIxX3QyLmh0bWwiIC8%2BZAIIEGRkFg4CBw8WAh8CBV9Zb3VyIHNlYXJjaCBvbiA8Yj5LZXl3b3JkczogUHl0aG9uOyBMb2NhdGlvbjogTG9uZG9uLCBTb3V0aCBFYXN0OyA8L2I%2BIHJldHVybmVkIDxiPjg1PC9iPiBqb2JzLmQCCQ8WAh4HVmlzaWJsZWhkAgsPFgIfAgUoVGhlIG1vc3QgcmVsZXZhbnQgam9icyBhcmUgbGlzdGVkIGZpcnN0LmQCEw8PFgIeC05hdmlnYXRlVXJsBQF%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%3D& Refinesearch%24txtKeywords=Python& Refinesearch%24txtLocation=London%2CSouth+East& Refinesearch%24ddlRadius=0& Refinesearch%24btnSearch=Search& ddlCompanyType=0& ddlSort=1'

    Read the article

  • Problem with XML Deserialization C#

    - by alex
    I am having trouble with XML deserialization. In a nutshell - I have 2 classes: SMSMessage SMSSendingResponse I call an API that takes a bunch of parameters (represented by SMSMessage class) It returns an XML response. The response looks like this: <?xml version="1.0" encoding="utf-8"?> <data> <status>1</status> <message>OK</message> <results> <result> <account>12345</account> <to>012345678</to> <from>054321</from> <message>Testing</message> <flash></flash> <replace></replace> <report></report> <concat></concat> <id>f8d3eea1cbf6771a4bb02af3fb15253e</id> </result> </results> </data> Here is the SMSMessage class (with the xml serialization attributes so far) using System.Xml.Serialization; namespace XMLSerializationHelp { [XmlRoot("results")] public class SMSMessage { public string To { get { return Result.To; } } public string From { get { return Result.From; } } public string Message { get { return Result.Message; } } [XmlElement("result")] public Result Result { get; set; } } } Here is SMSMessageSendingResponse: using System.Xml.Serialization; namespace XMLSerializationHelp { [XmlRoot("data")] public class SMSSendingResponse { //should come from the results/result/account element. in our example "12345" public string AccountNumber { get { return SMSMessage.Result.AccountNumber; } } //should come from the "status" xml element [XmlElement("status")] public string Status { get; set; } //should come from the "message" xml element (in our example - "OK") [XmlElement("message")] public string Message { get; set; } //should come from the "id" xml element (in our example - "f8d3eea1cbf6771a4bb02af3fb15253e") public string ResponseID { get { return SMSMessage.Result.ResponseID; } } //should be created from the results/result element - ignore flash, replace, report and concat elements for now. [XmlElement("results")] public SMSMessage SMSMessage { get; set; } } } Here is the other class (Result) - I want to get rid of this, so only the 2 previously mentioned classes remain using System.Xml.Serialization; namespace XMLSerializationHelp { [XmlRoot("result")] public class Result { [XmlElement("account")] public string AccountNumber{ get; set; } [XmlElement("to")] public string To { get; set; } [XmlElement("from")] public string From { get; set; } [XmlElement("message")] public string Message { get; set; } [XmlElement("id")] public string ResponseID { get; set; } } } I don't want SMSMessage to be aware of the SMSSendingResponse - as this will be handled by a different part of my application

    Read the article

  • Getting timing consistency in Linux

    - by Jim Hunziker
    I can't seem to get a simple program (with lots of memory access) to achieve consistent timing in Linux. I'm using a 2.6 kernel, and the program is being run on a dual-core processor with realtime priority. I'm trying to disable cache effects by declaring the memory arrays as volatile. Below are the results and the program. What are some possible sources of the outliers? Results: Number of trials: 100 Range: 0.021732s to 0.085596s Average Time: 0.058094s Standard Deviation: 0.006944s Extreme Outliers (2 SDs away from mean): 7 Average Time, excluding extreme outliers: 0.059273s Program: #include <stdio.h> #include <stdlib.h> #include <math.h> #include <sched.h> #include <sys/time.h> #define NUM_POINTS 5000000 #define REPS 100 unsigned long long getTimestamp() { unsigned long long usecCount; struct timeval timeVal; gettimeofday(&timeVal, 0); usecCount = timeVal.tv_sec * (unsigned long long) 1000000; usecCount += timeVal.tv_usec; return (usecCount); } double convertTimestampToSecs(unsigned long long timestamp) { return (timestamp / (double) 1000000); } int main(int argc, char* argv[]) { unsigned long long start, stop; double times[REPS]; double sum = 0; double scale, avg, newavg, median; double stddev = 0; double maxval = -1.0, minval = 1000000.0; int i, j, freq, count; int outliers = 0; struct sched_param sparam; sched_getparam(getpid(), &sparam); sparam.sched_priority = sched_get_priority_max(SCHED_FIFO); sched_setscheduler(getpid(), SCHED_FIFO, &sparam); volatile float* data; volatile float* results; data = calloc(NUM_POINTS, sizeof(float)); results = calloc(NUM_POINTS, sizeof(float)); for (i = 0; i < REPS; ++i) { start = getTimestamp(); for (j = 0; j < NUM_POINTS; ++j) { results[j] = data[j]; } stop = getTimestamp(); times[i] = convertTimestampToSecs(stop-start); } free(data); free(results); for (i = 0; i < REPS; i++) { sum += times[i]; if (times[i] > maxval) maxval = times[i]; if (times[i] < minval) minval = times[i]; } avg = sum/REPS; for (i = 0; i < REPS; i++) stddev += (times[i] - avg)*(times[i] - avg); stddev /= REPS; stddev = sqrt(stddev); for (i = 0; i < REPS; i++) { if (times[i] > avg + 2*stddev || times[i] < avg - 2*stddev) { sum -= times[i]; outliers++; } } newavg = sum/(REPS-outliers); printf("Number of trials: %d\n", REPS); printf("Range: %fs to %fs\n", minval, maxval); printf("Average Time: %fs\n", avg); printf("Standard Deviation: %fs\n", stddev); printf("Extreme Outliers (2 SDs away from mean): %d\n", outliers); printf("Average Time, excluding extreme outliers: %fs\n", newavg); return 0; }

    Read the article

  • Passing a variable from Excel 2007 Custom Task Pane to Hosted PowerShell

    - by Uros Calakovic
    I am testing PowerShell hosting using C#. Here is a console application that works: using System; using System.Collections; using System.Collections.Generic; using System.Collections.ObjectModel; using System.Management.Automation; using System.Management.Automation.Runspaces; using Microsoft.Office.Interop.Excel; namespace ConsoleApplication3 { class Program { static void Main() { Application app = new Application(); app.Visible = true; app.Workbooks.Add(XlWBATemplate.xlWBATWorksheet); Runspace runspace = RunspaceFactory.CreateRunspace(); runspace.Open(); runspace.SessionStateProxy.SetVariable("Application", app); Pipeline pipeline = runspace.CreatePipeline("$Application"); Collection<PSObject> results = null; try { results = pipeline.Invoke(); foreach (PSObject pob in results) { Console.WriteLine(pob); } } catch (RuntimeException re) { Console.WriteLine(re.GetType().Name); Console.WriteLine(re.Message); } } } } I first create an Excel.Application instance and pass it to the hosted PowerShell instance as a varible named $Application. This works and I can use this variable as if Excel.Application was created from within PowerShell. I next created an Excel addin using VS 2008 and added a user control with two text boxes and a button to the addin (the user control appears as a custom task pane when Excel starts). The idea was this: when I click the button a hosted PowerShell instance is created and I can pass to it the current Excel.Application instance as a variable, just like in the first sample, so I can use this variable to automate Excel from PowerShell (one text box would be used for input and the other one for output. Here is the code: using System; using System.Windows.Forms; using System.Management.Automation; using System.Management.Automation.Runspaces; using System.Collections.ObjectModel; using Microsoft.Office.Interop.Excel; namespace POSHAddin { public partial class POSHControl : UserControl { public POSHControl() { InitializeComponent(); } private void btnRun_Click(object sender, EventArgs e) { txtOutput.Clear(); Microsoft.Office.Interop.Excel.Application app = Globals.ThisAddIn.Application; Runspace runspace = RunspaceFactory.CreateRunspace(); runspace.Open(); runspace.SessionStateProxy.SetVariable("Application", app); Pipeline pipeline = runspace.CreatePipeline( "$Application | Get-Member | Out-String"); app.ActiveCell.Value2 = "Test"; Collection<PSObject> results = null; try { results = pipeline.Invoke(); foreach (PSObject pob in results) { txtOutput.Text += pob.ToString() + "-"; } } catch (RuntimeException re) { txtOutput.Text += re.GetType().Name; txtOutput.Text += re.Message; } } } } The code is similar to the first sample, except that the current Excel.Application instance is available to the addin via Globals.ThisAddIn.Application (VSTO generated) and I can see that it is really a Microsoft.Office.Interop.Excel.Application instance because I can use things like app.ActiveCell.Value2 = "Test" (this actually puts the text into the active cell). But when I pass the Excel.Application instance to the PowerShell instance what gets there is an instance of System.__ComObject and I can't figure out how to cast it to Excel.Application. When I examine the variable from PowerShell using $Application | Get-Member this is the output I get in the second text box: TypeName: System.__ComObject Name MemberType Definition ---- ---------- ---------- CreateObjRef Method System.Runtime.Remoting.ObjRef CreateObj... Equals Method System.Boolean Equals(Object obj) GetHashCode Method System.Int32 GetHashCode() GetLifetimeService Method System.Object GetLifetimeService() GetType Method System.Type GetType() InitializeLifetimeService Method System.Object InitializeLifetimeService() ToString Method System.String ToString() My question is how can I pass an instance of Microsoft.Office.Interop.Excel.Application from a VSTO generated Excel 2007 addin to a hosted PowerShell instance, so I can manipulate it from PowerShell? (I have previously posted the question in the Microsoft C# forum without an answer)

    Read the article

  • parallel computation for an Iterator of elements in Java

    - by Brian Harris
    I've had the same need a few times now and wanted to get other thoughts on the right way to structure a solution. The need is to perform some operation on many elements on many threads without needing to have all elements in memory at once, just the ones under computation. As in, Iterables.partition is insufficient because it brings all elements into memory up front. Expressing it in code, I want to write a BulkCalc2 that does the same thing as BulkCalc1, just in parallel. Below is sample code that illustrates my best attempt. I'm not satisfied because it's big and ugly, but it does seem to accomplish my goals of keeping threads highly utilized until the work is done, propagating any exceptions during computation, and not having more than numThreads instances of BigThing necessarily in memory at once. I'll accept the answer which meets the stated goals in the most concise way, whether it's a way to improve my BulkCalc2 or a completely different solution. interface BigThing { int getId(); String getString(); } class Calc { // somewhat expensive computation double calc(BigThing bigThing) { Random r = new Random(bigThing.getString().hashCode()); double d = 0; for (int i = 0; i < 100000; i++) { d += r.nextDouble(); } return d; } } class BulkCalc1 { final Calc calc; public BulkCalc1(Calc calc) { this.calc = calc; } public TreeMap<Integer, Double> calc(Iterator<BigThing> in) { TreeMap<Integer, Double> results = Maps.newTreeMap(); while (in.hasNext()) { BigThing o = in.next(); results.put(o.getId(), calc.calc(o)); } return results; } } class SafeIterator<T> { final Iterator<T> in; SafeIterator(Iterator<T> in) { this.in = in; } synchronized T nextOrNull() { if (in.hasNext()) { return in.next(); } return null; } } class BulkCalc2 { final Calc calc; final int numThreads; public BulkCalc2(Calc calc, int numThreads) { this.calc = calc; this.numThreads = numThreads; } public TreeMap<Integer, Double> calc(Iterator<BigThing> in) { ExecutorService e = Executors.newFixedThreadPool(numThreads); List<Future<?>> futures = Lists.newLinkedList(); final Map<Integer, Double> results = new MapMaker().concurrencyLevel(numThreads).makeMap(); final SafeIterator<BigThing> it = new SafeIterator<BigThing>(in); for (int i = 0; i < numThreads; i++) { futures.add(e.submit(new Runnable() { @Override public void run() { while (true) { BigThing o = it.nextOrNull(); if (o == null) { return; } results.put(o.getId(), calc.calc(o)); } } })); } e.shutdown(); for (Future<?> future : futures) { try { future.get(); } catch (InterruptedException ex) { // swallowing is OK } catch (ExecutionException ex) { throw Throwables.propagate(ex.getCause()); } } return new TreeMap<Integer, Double>(results); } }

    Read the article

  • How do you place an Excel Sheet/Workbook onto a C# .NET Winform?

    - by incognick
    I am trying to create a stand alone application in Visual Studio 2008 C# .Net that will house Excel Workbooks (2007). I am using Office.Interop in order to create the Excel application and I open the workbooks via Workbooks.Open(...). The Interop does not provide any functionality to "move" the workbooks onto a form so I turned to P/Invoke Win32 library. I am able to move the entire excel application onto a WinForm with great success: // pseudo code to give you the idea excel = new Excel.ApplicationClass(); SetParent(excel.Hwnd, form.handle); This allows me to customize the form and control user input. All right click commands and formula editing work properly. Now, the issue I run into is when I want to open two workbooks in two separate forms. I do this by creating two excel application classes and placing each of those in their own form. When I try to reference one workbook to another workbook via =[Book2]Sheet1!A1, for example, it does not update. This is expected as each application is running under its own thread/process. Here are the alternatives I have tried. If you have any suggestions I would be greatly appreciative.(OLE is not an option. VSTO must be available) Create a single application class and move the workbook window into my form. Results: The window moves into my form and displays correctly, however, no right click or left click works on the form and it never gains focus. (I have tried to manually set focus and it does not work either). My guess is, by moving the window outside of the XLDESK application (viewable in Spy++ for Excel Application), the workbook application (EXCEL7) does not receive the correct window messages to gain focus and to behave properly. This leads me to: Move the XLDESK window handle into my form. Results: This allows the workbook to be click-able again but also has an undesired result of moving all child windows into the same form. Create a main excel application that creates workbooks. Create a new excel application for each new window. Move the workbook under the new excel application XLDESK window. Results: This also has the same effect of the 1st option. Unable to click in the workbook. This must mean that the thread that created the workbook is also responsible for the events. Create a windows hook that watches the WndProc procedure. Results: No events watched. The targeted thread must export the hook proc in a DLL export call. Excel does not do this and thus you cannot inject into it's DLL (unless someone can prove me wrong). I am able to watch all threads within my own process but not from an outside process. Excel is created as a separate process. Subclass NativeWindow. Results: Same as #4. After I move the window into my form, all events are captured up until the mouse is directly over the excel sheet making the sheet seem unclickable. One idea I haven't tried yet is just to continually save the excel sheet as the user edits it. This should update all references but I would feel this would cause poor system performance. There will be numerous chart references as well and I'm not sure if this solution would cause problems further down the road. I think in the end, all the workbooks need to be created by the same Excel Application and then moved to get the desired results but I can't seem to find the correct way to move the windows without disabling the user input in the process. Any suggestions?

    Read the article

  • Python soap using soaplib (server) and suds (client)

    - by Celso Axelrud
    This question is related to: http://stackoverflow.com/questions/1751027/python-soap-server-client In the case of soap with python, there are recommendation to use soaplib (http://wiki.github.com/jkp/soaplib) as soap server and suds (https://fedorahosted.org/suds/) as soap client. My target is to create soap services in python that can be consumed by several clients (java, etc). I tried the HelloWorld example from soaplib (http://trac.optio.webfactional.com/wiki/HelloWorld). It works well when the client is also using soaplib. Then, I tried to use suds as client consuming the HelloWorld services and it fail. -Why this is happening? Does soaplib server has problems to consumed by different clients? Here the code for the server: from soaplib.wsgi_soap import SimpleWSGISoapApp from soaplib.service import soapmethod from soaplib.serializers.primitive import String, Integer, Arraycode class HelloWorldService(SimpleWSGISoapApp): @soapmethod(String,Integer,_returns=Array(String)) def say_hello(self,name,times): results = [] for i in range(0,times): results.append('Hello, %s'%name) return results if __name__=='__main__': from cherrypy.wsgiserver import CherryPyWSGIServer #from cherrypy._cpwsgiserver import CherryPyWSGIServer # this example uses CherryPy2.2, use cherrypy.wsgiserver.CherryPyWSGIServer for CherryPy 3.0 server = CherryPyWSGIServer(('localhost',7789),HelloWorldService()) server.start() This is the soaplib client: from soaplib.client import make_service_client from SoapServerTest_1 import HelloWorldService client = make_service_client('http://localhost:7789/',HelloWorldService()) print client.say_hello("Dave",5) Results: >>> ['Hello, Dave', 'Hello, Dave', 'Hello, Dave', 'Hello, Dave', 'Hello, Dave'] This is the suds client: from suds.client import Client url = 'http://localhost:7789/HelloWordService?wsdl' client1 = Client(url) client1.service.say_hello("Dave",5) Results: >>> Unhandled exception while debugging... Traceback (most recent call last): File "C:\Python25\Lib\site-packages\RTEP\Sequencing\SoapClientTest_1.py", line 10, in <module> client1.service.say_hello("Dave",5) File "c:\python25\lib\site-packages\suds\client.py", line 537, in __call__ return client.invoke(args, kwargs) File "c:\python25\lib\site-packages\suds\client.py", line 597, in invoke result = self.send(msg) File "c:\python25\lib\site-packages\suds\client.py", line 626, in send result = self.succeeded(binding, reply.message) File "c:\python25\lib\site-packages\suds\client.py", line 658, in succeeded r, p = binding.get_reply(self.method, reply) File "c:\python25\lib\site-packages\suds\bindings\binding.py", line 158, in get_reply result = unmarshaller.process(nodes[0], resolved) File "c:\python25\lib\site-packages\suds\umx\typed.py", line 66, in process return Core.process(self, content) File "c:\python25\lib\site-packages\suds\umx\core.py", line 48, in process return self.append(content) File "c:\python25\lib\site-packages\suds\umx\core.py", line 63, in append self.append_children(content) File "c:\python25\lib\site-packages\suds\umx\core.py", line 140, in append_children cval = self.append(cont) File "c:\python25\lib\site-packages\suds\umx\core.py", line 61, in append self.start(content) File "c:\python25\lib\site-packages\suds\umx\typed.py", line 77, in start found = self.resolver.find(content.node) File "c:\python25\lib\site-packages\suds\resolver.py", line 341, in find frame = Frame(result, resolved=known, ancestry=ancestry) File "c:\python25\lib\site-packages\suds\resolver.py", line 473, in __init__ resolved = type.resolve() File "c:\python25\lib\site-packages\suds\xsd\sxbasic.py", line 63, in resolve raise TypeNotFound(qref) TypeNotFound: Type not found: '(string, HelloWorldService.HelloWorldService, )'

    Read the article

  • Using FindAll on a List<List<T>> type

    - by Ken Foster
    Assuming public class MyClass { public int ID {get; set; } public string Name {get; set; } } and List<MyClass> classList = //populate with MyClass instances of various IDs I can do List<MyClass> result = classList.FindAll(class => class.ID == 123); and that will give me a list of just classes with ID = 123. Works great, looks elegant. Now, if I had List<List<MyClass>> listOfClassLists = //populate with Lists of MyClass instances How do I get a filtered list where the lists themselves are filtered. I tried List<List<MyClass>> result = listOfClassLists.FindAll (list => list.FindAll(class => class.ID == 123).Count > 0); it looks elegant, but doesn't work. It only includes Lists of classes where at least one class has an ID of 123, but it includes ALL MyClass instances in that list, not just the ones that match. I ended up having to do List<List<MyClass>> result = Results(listOfClassLists, 123); private List<List<MyClass>> Results(List<List<MyClass>> myListOfLists, int id) { List<List<MyClass>> results = new List<List<MyClass>>(); foreach (List<MyClass> myClassList in myListOfLists) { List<MyClass> subList = myClassList.FindAll(myClass => myClass.ID == id); if (subList.Count > 0) results.Add(subList); } return results; } which gets the job done, but isn't that elegant. Just looking for better ways to do a FindAll on a List of Lists. Ken

    Read the article

  • Javascript terminates after trying to select data from an object passed to a function

    - by Silmaril89
    Here is my javascript: $(document).ready(function(){ var queries = getUrlVars(); $.get("mail3.php", { listid: queries["listid"], mindex: queries["mindex"] }, showData, 'html'); }); function showData(data) { var response = $(data).find("#mailing").html(); if (response == null) { $("#results").html("<h3>Server didn't respond, try again.</h3>"); } else if (response.length) { var old = $("#results").html(); old = old + "<br /><h3>" + response + "</h3>"; $("#results").html(old); var words = response.split(' '); words[2] = words[2] * 1; words[4] = words[4] * 1; if (words[2] < words[4]) { var queries = getUrlVars(); $.get("mail3.php", { listid: queries["listid"], mindex: words[2] }, function(data){showData(data);}, 'html'); } else { var done = $(data).find("#done").html(); old = old + "<br />" + done; $("#results").html(old); } } else { $("#results").html("<h3>Server responded with an empty reply, try again.</h3>"); } } function getUrlVars() { var vars = [], hash; var hashes = window.location.href.slice(window.location.href.indexOf('?') + 1).split('&'); for (var i = 0; i < hashes.length; i++) { hash = hashes[i].split('='); vars.push(hash[0]); vars[hash[0]] = hash[1]; } return vars; } After the first line in showData: var response = $(data).find("#mailing").html(); the javascript stops. If I put an alert before it, the alert pops up, after it, it doesn't pop up. There must be something wrong with using $(data), but why? Any ideas would be appreciated.

    Read the article

  • How to perform gui operation in doInBackground method?

    - by jM2.me
    My application reads a user selected file which contains addresses and then displays on mapview when done geocoding. To avoid hanging app the importing and geocoding is done in AsyncTask. public class LoadOverlayAsync extends AsyncTask<Uri, Integer, StopsOverlay> { Context context; MapView mapView; Drawable drawable; public LoadOverlayAsync(Context con, MapView mv, Drawable dw) { context = con; mapView = mv; drawable = dw; } protected StopsOverlay doInBackground(Uri... uris) { StringBuilder text = new StringBuilder(); StopsOverlay stopsOverlay = new StopsOverlay(drawable, context); Geocoder geo = new Geocoder(context, Locale.US); try { File file = new File(new URI(uris[0].toString())); BufferedReader br = new BufferedReader(new FileReader(file)); String line; while ((line = br.readLine()) != null) { StopOverlay stopOverlay = null; String[] tempLine = line.split("~"); List<Address> results = geo.getFromLocationName(tempLine[4] + " " + tempLine[5] + " " + tempLine[7] + " " + tempLine[8], 10); if (results.size() > 0) { Toast progressToast = Toast.makeText(context, "More than one yo", 1000); progressToast.show(); } else if (results.size() == 1) { Address addr = results.get(0); GeoPoint mPoint = new GeoPoint((int)(addr.getLatitude() * 1E6), (int)(addr.getLongitude() * 1E6)); stopOverlay = new StopOverlay(mPoint, tempLine); } if (stopOverlay != null) { stopsOverlay.addOverlay(stopOverlay); } //List<Address> results = geo.getFromLocationName(locationName, maxResults) } } catch (URISyntaxException e) { showErrorToast(e.toString()); //e.printStackTrace(); } catch (FileNotFoundException e) { showErrorToast(e.toString()); //e.printStackTrace(); } catch (IOException e) { showErrorToast(e.toString()); //e.printStackTrace(); } return stopsOverlay; } protected void onProgressUpdate(Integer... progress) { Toast progressToast = Toast.makeText(context, "Loaded " + progress.toString(), 1000); progressToast.show(); } protected void onPostExecute(StopsOverlay so) { //mapView.getOverlays().add(so); Toast progressToast = Toast.makeText(context, "Done geocoding", 1000); progressToast.show(); } protected void showErrorToast(String msg) { Toast Newtoast = Toast.makeText(context, msg, 10000); Newtoast.show(); } } But if geocode fails, I want a dialog popup to let user edit the address. That would require calling on gui method while in doInBackground. What would be a good workaround this?

    Read the article

  • Is throwing an exception a healthy way to exit?

    - by ramaseshan
    I have a setup that looks like this. class Checker { // member data Results m_results; // see below public: bool Check(); private: bool Check1(); bool Check2(); // .. so on }; Checker is a class that performs lengthy check computations for engineering analysis. Each type of check has a resultant double that the checker stores. (see below) bool Checker::Check() { // initilisations etc. Check1(); Check2(); // ... so on } A typical Check function would look like this: bool Checker::Check1() { double result; // lots of code m_results.SetCheck1Result(result); } And the results class looks something like this: class Results { double m_check1Result; double m_check2Result; // ... public: void SetCheck1Result(double d); double GetOverallResult() { return max(m_check1Result, m_check2Result, ...); } }; Note: all code is oversimplified. The Checker and Result classes were initially written to perform all checks and return an overall double result. There is now a new requirement where I only need to know if any of the results exceeds 1. If it does, subsequent checks need not be carried out(it's an optimisation). To achieve this, I could either: Modify every CheckN function to keep check for result and return. The parent Check function would keep checking m_results. OR In the Results::SetCheckNResults(), throw an exception if the value exceeds 1 and catch it at the end of Checker::Check(). The first is tedious, error prone and sub-optimal because every CheckN function further branches out into sub-checks etc. The second is non-intrusive and quick. One disadvantage is I can think of is that the Checker code may not necessarily be exception-safe(although there is no other exception being thrown anywhere else). Is there anything else that's obvious that I'm overlooking? What about the cost of throwing exceptions and stack unwinding? Is there a better 3rd option?

    Read the article

  • Criticise/Recommendations for my code

    - by aLk
    Before i go any further it would be nice to know if there is any major design flaws in my program so far. Is there anything worth changing before i continue? Model package model; import java.sql.*; import java.util.*; public class MovieDatabase { @SuppressWarnings({ "rawtypes", "unchecked" }) public List queryMovies() throws SQLException { Connection connection = null; java.sql.Statement statement = null; ResultSet rs = null; List results = new ArrayList(); try { DriverManager.registerDriver(new com.mysql.jdbc.Driver()); connection = DriverManager.getConnection("jdbc:mysql://localhost:3306/test", "root", "password"); statement = connection.createStatement(); String query = "SELECT * FROM movie"; rs = statement.executeQuery(query); while(rs.next()) { MovieBean bean = new MovieBean(); bean.setMovieId(rs.getInt(1)); bean.setTitle(rs.getString(2)); bean.setYear(rs.getInt(3)); bean.setRating(rs.getInt(4)); results.add(bean); } } catch(SQLException e) { } return results; } } Servlet public class Service extends HttpServlet { @SuppressWarnings("rawtypes") protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException { response.setContentType("text/html"); PrintWriter out = response.getWriter(); out.println("Movies!"); MovieDatabase movies = new MovieDatabase(); try { List results = movies.queryMovies(); Iterator it = results.iterator(); while(it.hasNext()) { MovieBean movie = new MovieBean(); movie = (MovieBean)it.next(); out.println(movie.getYear()); } } catch(SQLException e) { } } } Bean package model; @SuppressWarnings("serial") public class MovieBean implements java.io.Serializable { protected int movieid; protected int rating; protected int year; protected String title; public MovieBean() { } public void setMovieId(int movieidVal) { movieid = movieidVal; } public void setRating(int ratingVal) { rating = ratingVal; } public void setYear(int yearVal) { year = yearVal; } public void setTitle(String titleVal) { title = titleVal; } public int getMovieId() { return movieid; } public int getRating() { return rating; } public int getYear() { return year; } public String getTitle() { return title; } }

    Read the article

  • Select Elements in nested Divs using JQuery

    - by PIKP
    I have the following html markup inside a Div named item and I want to select all the elements (inside nested divs) and clear the values. As shown in following given Jquery I have managed to access elements in each Div by using.children().each(). But the the problem is .children().each()goes one level down at a time from the parent div, so I have repeated the same code block with multiple .children() to access the elements inside nested Divs, can anyone suggest me a method to do this without repeating the code for N number of nested divs . html markup <div class="item"> <input type="hidden" value="1234" name="testVal"> <div class="form-group" id="c1"> <div class="controls "> <input type="text" value="Results" name="s1" maxlength="255" id="id2"> </div> </div> <div class="form-group" id="id4"> <input type="text" value="Results" name="s12" maxlength="255" id="id225"> <div class="form-group" id="id41"> <input type="text" value="Results" name="s12" maxlength="255" id="5"> <div class="form-group" id="id42"> <input type="text" value="Results" name="s12" maxlength="255" id="5"> <div class="form-group" id="id43"> <input type="text" value="Results" name="s12" maxlength="255" id="id224"> </div> </div> </div> </div> </div> My Qjuery script var row = $(".item:first").clone(false).get(0); $(row).children().each(function () { updateElementIndex(this, prefix, formCount); if ($(this).attr('type') == 'text') { $(this).val(''); } if ($(this).attr('type') == 'hidden' && ($(this).attr('name') != 'csrfmiddlewaretoken')) { $(this).val(''); } if ($(this).attr('type') == 'file') { $(this).val(''); } if ($(this).attr('type') == 'checkbox') { $(this).attr('checked', false); } $(this).remove('a'); }); // Relabel or rename all the relevant bits $(row).children().children().each(function () { updateElementIndex(this, prefix, formCount) if ($(this).attr('type') == 'text') { $(this).val(''); } if ($(this).attr('type') == 'hidden' && ($(this).attr('name') != 'csrfmiddlewaretoken')) { $(this).val(''); } if ($(this).attr('type') == 'file') { $(this).val(''); } if ($(this).attr('type') == 'checkbox') { $(this).attr('checked', false); } $(this).remove('a'); });

    Read the article

  • Node.js Adventure - Host Node.js on Windows Azure Worker Role

    - by Shaun
    In my previous post I demonstrated about how to develop and deploy a Node.js application on Windows Azure Web Site (a.k.a. WAWS). WAWS is a new feature in Windows Azure platform. Since it’s low-cost, and it provides IIS and IISNode components so that we can host our Node.js application though Git, FTP and WebMatrix without any configuration and component installation. But sometimes we need to use the Windows Azure Cloud Service (a.k.a. WACS) and host our Node.js on worker role. Below are some benefits of using worker role. - WAWS leverages IIS and IISNode to host Node.js application, which runs in x86 WOW mode. It reduces the performance comparing with x64 in some cases. - WACS worker role does not need IIS, hence there’s no restriction of IIS, such as 8000 concurrent requests limitation. - WACS provides more flexibility and controls to the developers. For example, we can RDP to the virtual machines of our worker role instances. - WACS provides the service configuration features which can be changed when the role is running. - WACS provides more scaling capability than WAWS. In WAWS we can have at most 3 reserved instances per web site while in WACS we can have up to 20 instances in a subscription. - Since when using WACS worker role we starts the node by ourselves in a process, we can control the input, output and error stream. We can also control the version of Node.js.   Run Node.js in Worker Role Node.js can be started by just having its execution file. This means in Windows Azure, we can have a worker role with the “node.exe” and the Node.js source files, then start it in Run method of the worker role entry class. Let’s create a new windows azure project in Visual Studio and add a new worker role. Since we need our worker role execute the “node.exe” with our application code we need to add the “node.exe” into our project. Right click on the worker role project and add an existing item. By default the Node.js will be installed in the “Program Files\nodejs” folder so we can navigate there and add the “node.exe”. Then we need to create the entry code of Node.js. In WAWS the entry file must be named “server.js”, which is because it’s hosted by IIS and IISNode and IISNode only accept “server.js”. But here as we control everything we can choose any files as the entry code. For example, I created a new JavaScript file named “index.js” in project root. Since we created a C# Windows Azure project we cannot create a JavaScript file from the context menu “Add new item”. We have to create a text file, and then rename it to JavaScript extension. After we added these two files we should set their “Copy to Output Directory” property to “Copy Always”, or “Copy if Newer”. Otherwise they will not be involved in the package when deployed. Let’s paste a very simple Node.js code in the “index.js” as below. As you can see I created a web server listening at port 12345. 1: var http = require("http"); 2: var port = 12345; 3:  4: http.createServer(function (req, res) { 5: res.writeHead(200, { "Content-Type": "text/plain" }); 6: res.end("Hello World\n"); 7: }).listen(port); 8:  9: console.log("Server running at port %d", port); Then we need to start “node.exe” with this file when our worker role was started. This can be done in its Run method. I found the Node.js and entry JavaScript file name, and then create a new process to run it. Our worker role will wait for the process to be exited. If everything is OK once our web server was opened the process will be there listening for incoming requests, and should not be terminated. The code in worker role would be like this. 1: public override void Run() 2: { 3: // This is a sample worker implementation. Replace with your logic. 4: Trace.WriteLine("NodejsHost entry point called", "Information"); 5:  6: // retrieve the node.exe and entry node.js source code file name. 7: var node = Environment.ExpandEnvironmentVariables(@"%RoleRoot%\approot\node.exe"); 8: var js = "index.js"; 9:  10: // prepare the process starting of node.exe 11: var info = new ProcessStartInfo(node, js) 12: { 13: CreateNoWindow = false, 14: ErrorDialog = true, 15: WindowStyle = ProcessWindowStyle.Normal, 16: UseShellExecute = false, 17: WorkingDirectory = Environment.ExpandEnvironmentVariables(@"%RoleRoot%\approot") 18: }; 19: Trace.WriteLine(string.Format("{0} {1}", node, js), "Information"); 20:  21: // start the node.exe with entry code and wait for exit 22: var process = Process.Start(info); 23: process.WaitForExit(); 24: } Then we can run it locally. In the computer emulator UI the worker role started and it executed the Node.js, then Node.js windows appeared. Open the browser to verify the website hosted by our worker role. Next let’s deploy it to azure. But we need some additional steps. First, we need to create an input endpoint. By default there’s no endpoint defined in a worker role. So we will open the role property window in Visual Studio, create a new input TCP endpoint to the port we want our website to use. In this case I will use 80. Even though we created a web server we should add a TCP endpoint of the worker role, since Node.js always listen on TCP instead of HTTP. And then changed the “index.js”, let our web server listen on 80. 1: var http = require("http"); 2: var port = 80; 3:  4: http.createServer(function (req, res) { 5: res.writeHead(200, { "Content-Type": "text/plain" }); 6: res.end("Hello World\n"); 7: }).listen(port); 8:  9: console.log("Server running at port %d", port); Then publish it to Windows Azure. And then in browser we can see our Node.js website was running on WACS worker role. We may encounter an error if we tried to run our Node.js website on 80 port at local emulator. This is because the compute emulator registered 80 and map the 80 endpoint to 81. But our Node.js cannot detect this operation. So when it tried to listen on 80 it will failed since 80 have been used.   Use NPM Modules When we are using WAWS to host Node.js, we can simply install modules we need, and then just publish or upload all files to WAWS. But if we are using WACS worker role, we have to do some extra steps to make the modules work. Assuming that we plan to use “express” in our application. Firstly of all we should download and install this module through NPM command. But after the install finished, they are just in the disk but not included in the worker role project. If we deploy the worker role right now the module will not be packaged and uploaded to azure. Hence we need to add them to the project. On solution explorer window click the “Show all files” button, select the “node_modules” folder and in the context menu select “Include In Project”. But that not enough. We also need to make all files in this module to “Copy always” or “Copy if newer”, so that they can be uploaded to azure with the “node.exe” and “index.js”. This is painful step since there might be many files in a module. So I created a small tool which can update a C# project file, make its all items as “Copy always”. The code is very simple. 1: static void Main(string[] args) 2: { 3: if (args.Length < 1) 4: { 5: Console.WriteLine("Usage: copyallalways [project file]"); 6: return; 7: } 8:  9: var proj = args[0]; 10: File.Copy(proj, string.Format("{0}.bak", proj)); 11:  12: var xml = new XmlDocument(); 13: xml.Load(proj); 14: var nsManager = new XmlNamespaceManager(xml.NameTable); 15: nsManager.AddNamespace("pf", "http://schemas.microsoft.com/developer/msbuild/2003"); 16:  17: // add the output setting to copy always 18: var contentNodes = xml.SelectNodes("//pf:Project/pf:ItemGroup/pf:Content", nsManager); 19: UpdateNodes(contentNodes, xml, nsManager); 20: var noneNodes = xml.SelectNodes("//pf:Project/pf:ItemGroup/pf:None", nsManager); 21: UpdateNodes(noneNodes, xml, nsManager); 22: xml.Save(proj); 23:  24: // remove the namespace attributes 25: var content = xml.InnerXml.Replace("<CopyToOutputDirectory xmlns=\"\">", "<CopyToOutputDirectory>"); 26: xml.LoadXml(content); 27: xml.Save(proj); 28: } 29:  30: static void UpdateNodes(XmlNodeList nodes, XmlDocument xml, XmlNamespaceManager nsManager) 31: { 32: foreach (XmlNode node in nodes) 33: { 34: var copyToOutputDirectoryNode = node.SelectSingleNode("pf:CopyToOutputDirectory", nsManager); 35: if (copyToOutputDirectoryNode == null) 36: { 37: var n = xml.CreateNode(XmlNodeType.Element, "CopyToOutputDirectory", null); 38: n.InnerText = "Always"; 39: node.AppendChild(n); 40: } 41: else 42: { 43: if (string.Compare(copyToOutputDirectoryNode.InnerText, "Always", true) != 0) 44: { 45: copyToOutputDirectoryNode.InnerText = "Always"; 46: } 47: } 48: } 49: } Please be careful when use this tool. I created only for demo so do not use it directly in a production environment. Unload the worker role project, execute this tool with the worker role project file name as the command line argument, it will set all items as “Copy always”. Then reload this worker role project. Now let’s change the “index.js” to use express. 1: var express = require("express"); 2: var app = express(); 3:  4: var port = 80; 5:  6: app.configure(function () { 7: }); 8:  9: app.get("/", function (req, res) { 10: res.send("Hello Node.js!"); 11: }); 12:  13: app.get("/User/:id", function (req, res) { 14: var id = req.params.id; 15: res.json({ 16: "id": id, 17: "name": "user " + id, 18: "company": "IGT" 19: }); 20: }); 21:  22: app.listen(port); Finally let’s publish it and have a look in browser.   Use Windows Azure SQL Database We can use Windows Azure SQL Database (a.k.a. WACD) from Node.js as well on worker role hosting. Since we can control the version of Node.js, here we can use x64 version of “node-sqlserver” now. This is better than if we host Node.js on WAWS since it only support x86. Just install the “node-sqlserver” module from NPM, copy the “sqlserver.node” from “Build\Release” folder to “Lib” folder. Include them in worker role project and run my tool to make them to “Copy always”. Finally update the “index.js” to use WASD. 1: var express = require("express"); 2: var sql = require("node-sqlserver"); 3:  4: var connectionString = "Driver={SQL Server Native Client 10.0};Server=tcp:{SERVER NAME}.database.windows.net,1433;Database={DATABASE NAME};Uid={LOGIN}@{SERVER NAME};Pwd={PASSWORD};Encrypt=yes;Connection Timeout=30;"; 5: var port = 80; 6:  7: var app = express(); 8:  9: app.configure(function () { 10: app.use(express.bodyParser()); 11: }); 12:  13: app.get("/", function (req, res) { 14: sql.open(connectionString, function (err, conn) { 15: if (err) { 16: console.log(err); 17: res.send(500, "Cannot open connection."); 18: } 19: else { 20: conn.queryRaw("SELECT * FROM [Resource]", function (err, results) { 21: if (err) { 22: console.log(err); 23: res.send(500, "Cannot retrieve records."); 24: } 25: else { 26: res.json(results); 27: } 28: }); 29: } 30: }); 31: }); 32:  33: app.get("/text/:key/:culture", function (req, res) { 34: sql.open(connectionString, function (err, conn) { 35: if (err) { 36: console.log(err); 37: res.send(500, "Cannot open connection."); 38: } 39: else { 40: var key = req.params.key; 41: var culture = req.params.culture; 42: var command = "SELECT * FROM [Resource] WHERE [Key] = '" + key + "' AND [Culture] = '" + culture + "'"; 43: conn.queryRaw(command, function (err, results) { 44: if (err) { 45: console.log(err); 46: res.send(500, "Cannot retrieve records."); 47: } 48: else { 49: res.json(results); 50: } 51: }); 52: } 53: }); 54: }); 55:  56: app.get("/sproc/:key/:culture", function (req, res) { 57: sql.open(connectionString, function (err, conn) { 58: if (err) { 59: console.log(err); 60: res.send(500, "Cannot open connection."); 61: } 62: else { 63: var key = req.params.key; 64: var culture = req.params.culture; 65: var command = "EXEC GetItem '" + key + "', '" + culture + "'"; 66: conn.queryRaw(command, function (err, results) { 67: if (err) { 68: console.log(err); 69: res.send(500, "Cannot retrieve records."); 70: } 71: else { 72: res.json(results); 73: } 74: }); 75: } 76: }); 77: }); 78:  79: app.post("/new", function (req, res) { 80: var key = req.body.key; 81: var culture = req.body.culture; 82: var val = req.body.val; 83:  84: sql.open(connectionString, function (err, conn) { 85: if (err) { 86: console.log(err); 87: res.send(500, "Cannot open connection."); 88: } 89: else { 90: var command = "INSERT INTO [Resource] VALUES ('" + key + "', '" + culture + "', N'" + val + "')"; 91: conn.queryRaw(command, function (err, results) { 92: if (err) { 93: console.log(err); 94: res.send(500, "Cannot retrieve records."); 95: } 96: else { 97: res.send(200, "Inserted Successful"); 98: } 99: }); 100: } 101: }); 102: }); 103:  104: app.listen(port); Publish to azure and now we can see our Node.js is working with WASD through x64 version “node-sqlserver”.   Summary In this post I demonstrated how to host our Node.js in Windows Azure Cloud Service worker role. By using worker role we can control the version of Node.js, as well as the entry code. And it’s possible to do some pre jobs before the Node.js application started. It also removed the IIS and IISNode limitation. I personally recommended to use worker role as our Node.js hosting. But there are some problem if you use the approach I mentioned here. The first one is, we need to set all JavaScript files and module files as “Copy always” or “Copy if newer” manually. The second one is, in this way we cannot retrieve the cloud service configuration information. For example, we defined the endpoint in worker role property but we also specified the listening port in Node.js hardcoded. It should be changed that our Node.js can retrieve the endpoint. But I can tell you it won’t be working here. In the next post I will describe another way to execute the “node.exe” and Node.js application, so that we can get the cloud service configuration in Node.js. I will also demonstrate how to use Windows Azure Storage from Node.js by using the Windows Azure Node.js SDK.   Hope this helps, Shaun All documents and related graphics, codes are provided "AS IS" without warranty of any kind. Copyright © Shaun Ziyan Xu. This work is licensed under the Creative Commons License.

    Read the article

  • How to find and fix performance problems in ORM powered applications

    - by FransBouma
    Once in a while we get requests about how to fix performance problems with our framework. As it comes down to following the same steps and looking into the same things every single time, I decided to write a blogpost about it instead, so more people can learn from this and solve performance problems in their O/R mapper powered applications. In some parts it's focused on LLBLGen Pro but it's also usable for other O/R mapping frameworks, as the vast majority of performance problems in O/R mapper powered applications are not specific for a certain O/R mapper framework. Too often, the developer looks at the wrong part of the application, trying to fix what isn't a problem in that part, and getting frustrated that 'things are so slow with <insert your favorite framework X here>'. I'm in the O/R mapper business for a long time now (almost 10 years, full time) and as it's a small world, we O/R mapper developers know almost all tricks to pull off by now: we all know what to do to make task ABC faster and what compromises (because there are almost always compromises) to deal with if we decide to make ABC faster that way. Some O/R mapper frameworks are faster in X, others in Y, but you can be sure the difference is mainly a result of a compromise some developers are willing to deal with and others aren't. That's why the O/R mapper frameworks on the market today are different in many ways, even though they all fetch and save entities from and to a database. I'm not suggesting there's no room for improvement in today's O/R mapper frameworks, there always is, but it's not a matter of 'the slowness of the application is caused by the O/R mapper' anymore. Perhaps query generation can be optimized a bit here, row materialization can be optimized a bit there, but it's mainly coming down to milliseconds. Still worth it if you're a framework developer, but it's not much compared to the time spend inside databases and in user code: if a complete fetch takes 40ms or 50ms (from call to entity object collection), it won't make a difference for your application as that 10ms difference won't be noticed. That's why it's very important to find the real locations of the problems so developers can fix them properly and don't get frustrated because their quest to get a fast, performing application failed. Performance tuning basics and rules Finding and fixing performance problems in any application is a strict procedure with four prescribed steps: isolate, analyze, interpret and fix, in that order. It's key that you don't skip a step nor make assumptions: these steps help you find the reason of a problem which seems to be there, and how to fix it or leave it as-is. Skipping a step, or when you assume things will be bad/slow without doing analysis will lead to the path of premature optimization and won't actually solve your problems, only create new ones. The most important rule of finding and fixing performance problems in software is that you have to understand what 'performance problem' actually means. Most developers will say "when a piece of software / code is slow, you have a performance problem". But is that actually the case? If I write a Linq query which will aggregate, group and sort 5 million rows from several tables to produce a resultset of 10 rows, it might take more than a couple of milliseconds before that resultset is ready to be consumed by other logic. If I solely look at the Linq query, the code consuming the resultset of the 10 rows and then look at the time it takes to complete the whole procedure, it will appear to me to be slow: all that time taken to produce and consume 10 rows? But if you look closer, if you analyze and interpret the situation, you'll see it does a tremendous amount of work, and in that light it might even be extremely fast. With every performance problem you encounter, always do realize that what you're trying to solve is perhaps not a technical problem at all, but a perception problem. The second most important rule you have to understand is based on the old saying "Penny wise, Pound Foolish": the part which takes e.g. 5% of the total time T for a given task isn't worth optimizing if you have another part which takes a much larger part of the total time T for that same given task. Optimizing parts which are relatively insignificant for the total time taken is not going to bring you better results overall, even if you totally optimize that part away. This is the core reason why analysis of the complete set of application parts which participate in a given task is key to being successful in solving performance problems: No analysis -> no problem -> no solution. One warning up front: hunting for performance will always include making compromises. Fast software can be made maintainable, but if you want to squeeze as much performance out of your software, you will inevitably be faced with the dilemma of compromising one or more from the group {readability, maintainability, features} for the extra performance you think you'll gain. It's then up to you to decide whether it's worth it. In almost all cases it's not. The reason for this is simple: the vast majority of performance problems can be solved by implementing the proper algorithms, the ones with proven Big O-characteristics so you know the performance you'll get plus you know the algorithm will work. The time taken by the algorithm implementing code is inevitable: you already implemented the best algorithm. You might find some optimizations on the technical level but in general these are minor. Let's look at the four steps to see how they guide us through the quest to find and fix performance problems. Isolate The first thing you need to do is to isolate the areas in your application which are assumed to be slow. For example, if your application is a web application and a given page is taking several seconds or even minutes to load, it's a good candidate to check out. It's important to start with the isolate step because it allows you to focus on a single code path per area with a clear begin and end and ignore the rest. The rest of the steps are taken per identified problematic area. Keep in mind that isolation focuses on tasks in an application, not code snippets. A task is something that's started in your application by either another task or the user, or another program, and has a beginning and an end. You can see a task as a piece of functionality offered by your application.  Analyze Once you've determined the problem areas, you have to perform analysis on the code paths of each area, to see where the performance problems occur and which areas are not the problem. This is a multi-layered effort: an application which uses an O/R mapper typically consists of multiple parts: there's likely some kind of interface (web, webservice, windows etc.), a part which controls the interface and business logic, the O/R mapper part and the RDBMS, all connected with either a network or inter-process connections provided by the OS or other means. Each of these parts, including the connectivity plumbing, eat up a part of the total time it takes to complete a task, e.g. load a webpage with all orders of a given customer X. To understand which parts participate in the task / area we're investigating and how much they contribute to the total time taken to complete the task, analysis of each participating task is essential. Start with the code you wrote which starts the task, analyze the code and track the path it follows through your application. What does the code do along the way, verify whether it's correct or not. Analyze whether you have implemented the right algorithms in your code for this particular area. Remember we're looking at one area at a time, which means we're ignoring all other code paths, just the code path of the current problematic area, from begin to end and back. Don't dig in and start optimizing at the code level just yet. We're just analyzing. If your analysis reveals big architectural stupidity, it's perhaps a good idea to rethink the architecture at this point. For the rest, we're analyzing which means we collect data about what could be wrong, for each participating part of the complete application. Reviewing the code you wrote is a good tool to get deeper understanding of what is going on for a given task but ultimately it lacks precision and overview what really happens: humans aren't good code interpreters, computers are. We therefore need to utilize tools to get deeper understanding about which parts contribute how much time to the total task, triggered by which other parts and for example how many times are they called. There are two different kind of tools which are necessary: .NET profilers and O/R mapper / RDBMS profilers. .NET profiling .NET profilers (e.g. dotTrace by JetBrains or Ants by Red Gate software) show exactly which pieces of code are called, how many times they're called, and the time it took to run that piece of code, at the method level and sometimes even at the line level. The .NET profilers are essential tools for understanding whether the time taken to complete a given task / area in your application is consumed by .NET code, where exactly in your code, the path to that code, how many times that code was called by other code and thus reveals where hotspots are located: the areas where a solution can be found. Importantly, they also reveal which areas can be left alone: remember our penny wise pound foolish saying: if a profiler reveals that a group of methods are fast, or don't contribute much to the total time taken for a given task, ignore them. Even if the code in them is perhaps complex and looks like a candidate for optimization: you can work all day on that, it won't matter.  As we're focusing on a single area of the application, it's best to start profiling right before you actually activate the task/area. Most .NET profilers support this by starting the application without starting the profiling procedure just yet. You navigate to the particular part which is slow, start profiling in the profiler, in your application you perform the actions which are considered slow, and afterwards you get a snapshot in the profiler. The snapshot contains the data collected by the profiler during the slow action, so most data is produced by code in the area to investigate. This is important, because it allows you to stay focused on a single area. O/R mapper and RDBMS profiling .NET profilers give you a good insight in the .NET side of things, but not in the RDBMS side of the application. As this article is about O/R mapper powered applications, we're also looking at databases, and the software making it possible to consume the database in your application: the O/R mapper. To understand which parts of the O/R mapper and database participate how much to the total time taken for task T, we need different tools. There are two kind of tools focusing on O/R mappers and database performance profiling: O/R mapper profilers and RDBMS profilers. For O/R mapper profilers, you can look at LLBLGen Prof by hibernating rhinos or the Linq to Sql/LLBLGen Pro profiler by Huagati. Hibernating rhinos also have profilers for other O/R mappers like NHibernate (NHProf) and Entity Framework (EFProf) and work the same as LLBLGen Prof. For RDBMS profilers, you have to look whether the RDBMS vendor has a profiler. For example for SQL Server, the profiler is shipped with SQL Server, for Oracle it's build into the RDBMS, however there are also 3rd party tools. Which tool you're using isn't really important, what's important is that you get insight in which queries are executed during the task / area we're currently focused on and how long they took. Here, the O/R mapper profilers have an advantage as they collect the time it took to execute the query from the application's perspective so they also collect the time it took to transport data across the network. This is important because a query which returns a massive resultset or a resultset with large blob/clob/ntext/image fields takes more time to get transported across the network than a small resultset and a database profiler doesn't take this into account most of the time. Another tool to use in this case, which is more low level and not all O/R mappers support it (though LLBLGen Pro and NHibernate as well do) is tracing: most O/R mappers offer some form of tracing or logging system which you can use to collect the SQL generated and executed and often also other activity behind the scenes. While tracing can produce a tremendous amount of data in some cases, it also gives insight in what's going on. Interpret After we've completed the analysis step it's time to look at the data we've collected. We've done code reviews to see whether we've done anything stupid and which parts actually take place and if the proper algorithms have been implemented. We've done .NET profiling to see which parts are choke points and how much time they contribute to the total time taken to complete the task we're investigating. We've performed O/R mapper profiling and RDBMS profiling to see which queries were executed during the task, how many queries were generated and executed and how long they took to complete, including network transportation. All this data reveals two things: which parts are big contributors to the total time taken and which parts are irrelevant. Both aspects are very important. The parts which are irrelevant (i.e. don't contribute significantly to the total time taken) can be ignored from now on, we won't look at them. The parts which contribute a lot to the total time taken are important to look at. We now have to first look at the .NET profiler results, to see whether the time taken is consumed in our own code, in .NET framework code, in the O/R mapper itself or somewhere else. For example if most of the time is consumed by DbCommand.ExecuteReader, the time it took to complete the task is depending on the time the data is fetched from the database. If there was just 1 query executed, according to tracing or O/R mapper profilers / RDBMS profilers, check whether that query is optimal, uses indexes or has to deal with a lot of data. Interpret means that you follow the path from begin to end through the data collected and determine where, along the path, the most time is contributed. It also means that you have to check whether this was expected or is totally unexpected. My previous example of the 10 row resultset of a query which groups millions of rows will likely reveal that a long time is spend inside the database and almost no time is spend in the .NET code, meaning the RDBMS part contributes the most to the total time taken, the rest is compared to that time, irrelevant. Considering the vastness of the source data set, it's expected this will take some time. However, does it need tweaking? Perhaps all possible tweaks are already in place. In the interpret step you then have to decide that further action in this area is necessary or not, based on what the analysis results show: if the analysis results were unexpected and in the area where the most time is contributed to the total time taken is room for improvement, action should be taken. If not, you can only accept the situation and move on. In all cases, document your decision together with the analysis you've done. If you decide that the perceived performance problem is actually expected due to the nature of the task performed, it's essential that in the future when someone else looks at the application and starts asking questions you can answer them properly and new analysis is only necessary if situations changed. Fix After interpreting the analysis results you've concluded that some areas need adjustment. This is the fix step: you're actively correcting the performance problem with proper action targeted at the real cause. In many cases related to O/R mapper powered applications it means you'll use different features of the O/R mapper to achieve the same goal, or apply optimizations at the RDBMS level. It could also mean you apply caching inside your application (compromise memory consumption over performance) to avoid unnecessary re-querying data and re-consuming the results. After applying a change, it's key you re-do the analysis and interpretation steps: compare the results and expectations with what you had before, to see whether your actions had any effect or whether it moved the problem to a different part of the application. Don't fall into the trap to do partly analysis: do the full analysis again: .NET profiling and O/R mapper / RDBMS profiling. It might very well be that the changes you've made make one part faster but another part significantly slower, in such a way that the overall problem hasn't changed at all. Performance tuning is dealing with compromises and making choices: to use one feature over the other, to accept a higher memory footprint, to go away from the strict-OO path and execute queries directly onto the RDBMS, these are choices and compromises which will cross your path if you want to fix performance problems with respect to O/R mappers or data-access and databases in general. In most cases it's not a big issue: alternatives are often good choices too and the compromises aren't that hard to deal with. What is important is that you document why you made a choice, a compromise: which analysis data, which interpretation led you to the choice made. This is key for good maintainability in the years to come. Most common performance problems with O/R mappers Below is an incomplete list of common performance problems related to data-access / O/R mappers / RDBMS code. It will help you with fixing the hotspots you found in the interpretation step. SELECT N+1: (Lazy-loading specific). Lazy loading triggered performance bottlenecks. Consider a list of Orders bound to a grid. You have a Field mapped onto a related field in Order, Customer.CompanyName. Showing this column in the grid will make the grid fetch (indirectly) for each row the Customer row. This means you'll get for the single list not 1 query (for the orders) but 1+(the number of orders shown) queries. To solve this: use eager loading using a prefetch path to fetch the customers with the orders. SELECT N+1 is easy to spot with an O/R mapper profiler or RDBMS profiler: if you see a lot of identical queries executed at once, you have this problem. Prefetch paths using many path nodes or sorting, or limiting. Eager loading problem. Prefetch paths can help with performance, but as 1 query is fetched per node, it can be the number of data fetched in a child node is bigger than you think. Also consider that data in every node is merged on the client within the parent. This is fast, but it also can take some time if you fetch massive amounts of entities. If you keep fetches small, you can use tuning parameters like the ParameterizedPrefetchPathThreshold setting to get more optimal queries. Deep inheritance hierarchies of type Target Per Entity/Type. If you use inheritance of type Target per Entity / Type (each type in the inheritance hierarchy is mapped onto its own table/view), fetches will join subtype- and supertype tables in many cases, which can lead to a lot of performance problems if the hierarchy has many types. With this problem, keep inheritance to a minimum if possible, or switch to a hierarchy of type Target Per Hierarchy, which means all entities in the inheritance hierarchy are mapped onto the same table/view. Of course this has its own set of drawbacks, but it's a compromise you might want to take. Fetching massive amounts of data by fetching large lists of entities. LLBLGen Pro supports paging (and limiting the # of rows returned), which is often key to process through large sets of data. Use paging on the RDBMS if possible (so a query is executed which returns only the rows in the page requested). When using paging in a web application, be sure that you switch server-side paging on on the datasourcecontrol used. In this case, paging on the grid alone is not enough: this can lead to fetching a lot of data which is then loaded into the grid and paged there. Keep note that analyzing queries for paging could lead to the false assumption that paging doesn't occur, e.g. when the query contains a field of type ntext/image/clob/blob and DISTINCT can't be applied while it should have (e.g. due to a join): the datareader will do DISTINCT filtering on the client. this is a little slower but it does perform paging functionality on the data-reader so it won't fetch all rows even if the query suggests it does. Fetch massive amounts of data because blob/clob/ntext/image fields aren't excluded. LLBLGen Pro supports field exclusion for queries. You can exclude fields (also in prefetch paths) per query to avoid fetching all fields of an entity, e.g. when you don't need them for the logic consuming the resultset. Excluding fields can greatly reduce the amount of time spend on data-transport across the network. Use this optimization if you see that there's a big difference between query execution time on the RDBMS and the time reported by the .NET profiler for the ExecuteReader method call. Doing client-side aggregates/scalar calculations by consuming a lot of data. If possible, try to formulate a scalar query or group by query using the projection system or GetScalar functionality of LLBLGen Pro to do data consumption on the RDBMS server. It's far more efficient to process data on the RDBMS server than to first load it all in memory, then traverse the data in-memory to calculate a value. Using .ToList() constructs inside linq queries. It might be you use .ToList() somewhere in a Linq query which makes the query be run partially in-memory. Example: var q = from c in metaData.Customers.ToList() where c.Country=="Norway" select c; This will actually fetch all customers in-memory and do an in-memory filtering, as the linq query is defined on an IEnumerable<T>, and not on the IQueryable<T>. Linq is nice, but it can often be a bit unclear where some parts of a Linq query might run. Fetching all entities to delete into memory first. To delete a set of entities it's rather inefficient to first fetch them all into memory and then delete them one by one. It's more efficient to execute a DELETE FROM ... WHERE query on the database directly to delete the entities in one go. LLBLGen Pro supports this feature, and so do some other O/R mappers. It's not always possible to do this operation in the context of an O/R mapper however: if an O/R mapper relies on a cache, these kind of operations are likely not supported because they make it impossible to track whether an entity is actually removed from the DB and thus can be removed from the cache. Fetching all entities to update with an expression into memory first. Similar to the previous point: it is more efficient to update a set of entities directly with a single UPDATE query using an expression instead of fetching the entities into memory first and then updating the entities in a loop, and afterwards saving them. It might however be a compromise you don't want to take as it is working around the idea of having an object graph in memory which is manipulated and instead makes the code fully aware there's a RDBMS somewhere. Conclusion Performance tuning is almost always about compromises and making choices. It's also about knowing where to look and how the systems in play behave and should behave. The four steps I provided should help you stay focused on the real problem and lead you towards the solution. Knowing how to optimally use the systems participating in your own code (.NET framework, O/R mapper, RDBMS, network/services) is key for success as well as knowing what's going on inside the application you built. I hope you'll find this guide useful in tracking down performance problems and dealing with them in a useful way.  

    Read the article

< Previous Page | 68 69 70 71 72 73 74 75 76 77 78 79  | Next Page >