Search Results

Search found 2909 results on 117 pages for 'gang of four'.

Page 41/117 | < Previous Page | 37 38 39 40 41 42 43 44 45 46 47 48  | Next Page >

• linux raid 1: right after replacing and syncing one drive, the other disk fails - understanding what is going on with mdstat/mdadm

  - by devicerandom

  We have an old RAID 1 Linux server (Ubuntu Lucid 10.04), with four partitions. A few days ago /dev/sdb failed, and today we noticed /dev/sda had pre-failure ominous SMART signs (~4000 reallocated sector count). We replaced /dev/sdb this morning and rebuilt the RAID on the new drive, following this guide: http://www.howtoforge.com/replacing_hard_disks_in_a_raid1_array Everything went smooth until the very end. When it looked like it was finishing to synchronize the last partition, the other old one failed. At this point I am very unsure of the state of the system. Everything seems working and the files seem to be all accessible, just as if it synchronized everything, but I'm new to RAID and I'm worried about what is going on. The /proc/mdstat output is: Personalities : [raid1] [linear] [multipath] [raid0] [raid6] [raid5] [raid4] [raid10] md3 : active raid1 sdb4[2](S) sda4[0] 478713792 blocks [2/1] [U_] md2 : active raid1 sdb3[1] sda3[2](F) 244140992 blocks [2/1] [_U] md1 : active raid1 sdb2[1] sda2[2](F) 244140992 blocks [2/1] [_U] md0 : active raid1 sdb1[1] sda1[2](F) 9764800 blocks [2/1] [_U] unused devices: <none> The order of [_U] vs [U_]. Why aren't they consistent along all the array? Is the first U /dev/sda or /dev/sdb? (I tried looking on the web for this trivial information but I found no explicit indication) If I read correctly for md0, [_U] should be /dev/sda1 (down) and /dev/sdb1 (up). But if /dev/sda has failed, how can it be the opposite for md3 ? I understand /dev/sdb4 is now spare because probably it failed to synchronize it 100%, but why does it show /dev/sda4 as up? Shouldn't it be [__]? Or [_U] anyway? The /dev/sda drive now cannot even be accessed by SMART anymore apparently, so I wouldn't expect it to be up. What is wrong with my interpretation of the output? I attach also the outputs of mdadm --detail for the four partitions: /dev/md0: Version : 00.90 Creation Time : Fri Jan 21 18:43:07 2011 Raid Level : raid1 Array Size : 9764800 (9.31 GiB 10.00 GB) Used Dev Size : 9764800 (9.31 GiB 10.00 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 0 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:27:33 2013 State : clean, degraded Active Devices : 1 Working Devices : 1 Failed Devices : 1 Spare Devices : 0 UUID : a3b4dbbd:859bf7f2:bde36644:fcef85e2 Events : 0.7704 Number Major Minor RaidDevice State 0 0 0 0 removed 1 8 17 1 active sync /dev/sdb1 2 8 1 - faulty spare /dev/sda1 /dev/md1: Version : 00.90 Creation Time : Fri Jan 21 18:43:15 2011 Raid Level : raid1 Array Size : 244140992 (232.83 GiB 250.00 GB) Used Dev Size : 244140992 (232.83 GiB 250.00 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 1 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:39:06 2013 State : clean, degraded Active Devices : 1 Working Devices : 1 Failed Devices : 1 Spare Devices : 0 UUID : 8bcd5765:90dc93d5:cc70849c:224ced45 Events : 0.1508280 Number Major Minor RaidDevice State 0 0 0 0 removed 1 8 18 1 active sync /dev/sdb2 2 8 2 - faulty spare /dev/sda2 /dev/md2: Version : 00.90 Creation Time : Fri Jan 21 18:43:19 2011 Raid Level : raid1 Array Size : 244140992 (232.83 GiB 250.00 GB) Used Dev Size : 244140992 (232.83 GiB 250.00 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 2 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:46:44 2013 State : clean, degraded Active Devices : 1 Working Devices : 1 Failed Devices : 1 Spare Devices : 0 UUID : 2885668b:881cafed:b8275ae8:16bc7171 Events : 0.2289636 Number Major Minor RaidDevice State 0 0 0 0 removed 1 8 19 1 active sync /dev/sdb3 2 8 3 - faulty spare /dev/sda3 /dev/md3: Version : 00.90 Creation Time : Fri Jan 21 18:43:22 2011 Raid Level : raid1 Array Size : 478713792 (456.54 GiB 490.20 GB) Used Dev Size : 478713792 (456.54 GiB 490.20 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 3 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:19:20 2013 State : clean, degraded Active Devices : 1 Working Devices : 2 Failed Devices : 0 Spare Devices : 1 Number Major Minor RaidDevice State 0 8 4 0 active sync /dev/sda4 1 0 0 1 removed 2 8 20 - spare /dev/sdb4 The active sync on /dev/sda4 baffles me. I am worried because if tomorrow morning I have to replace /dev/sda, I want to be sure what should I sync with what and what is going on. I am also quite baffled by the fact /dev/sda decided to fail exactly when the raid finished resyncing. I'd like to understand what is really happening. Thanks a lot for your patience and help. Massimo

  Read the article

• jQuery urlencode/decode patch help

  - by jeerose

  Hi Gang, I'm using this jQuery urlencode and urldecode plugin - very simple and easy to use but it doesn't, in its original form, remove + from the string. The one comment on the home page suggests a patch but I don't know how to implement it. Can anyone help me out? The Page: http://www.digitalbart.com/jquery-and-urlencode/ //URL Encode/Decode $.extend({URLEncode:function(c){var o='';var x=0;c=c.toString(); var r=/(^[a-zA-Z0-9_.]*)/; while(x<c.length){var m=r.exec(c.substr(x)); if(m!=null && m.length>1 && m[1]!=''){o+=m[1];x+=m[1].length; }else{if(c[x]==' ')o+='+';else{var d=c.charCodeAt(x);var h=d.toString(16); o+='%'+(h.length<2?'0':'')+h.toUpperCase();}x++;}}return o;}, URLDecode:function(s){var o=s;var binVal,t;var r=/(%[^%]{2})/; while((m=r.exec(o))!=null && m.length>1 && m[1]!=''){ b=parseInt(m[1].substr(1),16); t=String.fromCharCode(b);o=o.replace(m[1],t);}return o;} }); The proposed Patch: function dummy_url_decode(url) { // fixed -- + char decodes to space char var o = url; var binVal, t, b; var r = /(%[^%]{2}|\+)/; while ((m = r.exec(o)) != null && m.length > 1 && m[1] != '') { if (m[1] == '+') { t = ' '; } else { b = parseInt(m[1].substr(1), 16); t = String.fromCharCode(b); } o = o.replace(m[1], t); } return o; } Thanks!

  Read the article

• Using htaccess rewrite/redirect so single PHP file can display data according to GET/POST variables

  - by revive

  Hey gang, Bear with me as I try to learn more about .htaccess redirect rules - I'm a UI guy by profession and I'm doing my best to enhance my coding skills languages other than HTML/CSS/PHP, etc So, what I have is an index.php file that contains a menu - pretty simple. If Javascript is enabled on the users computer then forms are shown below the menu using simple jQuery goodies, with each link having a 'return false;' applied. What I am trying to do is make the page accessible with JS turned off, so instead of redirecting the user to a different page, I would like to use POST or GET variables from each link, to show the various forms.. within the same index.php file - but have the URL reflect the menu choice Here is an example using menu items named - Home, About, Form1 and Form2: With JS on, clicking on of the above example buttons simply slides the form or containing the data down onto the page. With JS off, if the user clicks the 'About' link, I would like to rewrite the URL to be http://domain.com/about - but keep the user on the index.php page (since it is the only page) and be able to use POST or GET variable to show the using PHP. So, accessing http://domain.com/index.php?page=about would show http://domain.com/about in the URL but pass the GET variable 'page' to the index.php file. Hope that makes sense.. and I'm sure there are many ways to accomplish this, so if you have ideas of how to improve on it, by all means, I'd love to hear it ! Thanks again Revive

  Read the article

• Patterns for non-layered applications

  - by Paul Stovell

  In Patterns of Enterprise Application Architecture, Martin Fowler writes: This book is thus about how you decompose an enterprise application into layers and how those layers work together. Most nontrivial enterprise applications use a layered architecture of some form, but in some situations other approaches, such as pipes and filters, are valuable. I don't go into those situations, focussing instead on the context of a layered architecture because it's the most widely useful. What patterns exist for building non-layered applications/parts of an application? Take a statistical modelling engine for a financial institution. There might be a layer for data access, but I expect that most of the code would be in a single layer. Would you still expect to see Gang of Four patterns in such a layer? How about a domain model? Would you use OO at all, or would it be purely functional? The quote mentions pipes and filters as alternate models to layers. I can easily imagine a such an engine using pipes as a way to break down the data processing. What other patterns exist? Are there common patterns for areas like task scheduling, results aggregation, or work distribution? What are some alternatives to MapReduce?

  Read the article

• Design patterns for Agent / Actor based concurrent design.

  - by nso1

  Recently i have been getting into alternative languages that support an actor/agent/shared nothing architecture - ie. scala, clojure etc (clojure also supports shared state). So far most of the documentation that I have read focus around the intro level. What I am looking for is more advanced documentation along the gang of four but instead shared nothing based. Why ? It helps to grok the change in design thinking. Simple examples are easy, but in a real world java application (single threaded) you can have object graphs with 1000's of members with complex relationships. But with agent based concurrency development it introduces a whole new set of ideas to comprehend when designing large systems. ie. Agent granularity - how much state should one agent manage - implications on performance etc or are their good patterns for mapping shared state object graphs to agent based system. tips on mapping domain models to design. Discussions not on the technology but more on how to BEST use the technology in design (real world "complex" examples would be great).

  Read the article

• Loading .dll/.exe from file into temporary AppDomain throws Exception

  Hi Gang, I am trying to make a Visual Studio AddIn that removes unused references from the projects in the current solution (I know this can be done, Resharper does it, but my client doesn't want to pay for 300 licences). Anyhoo, I use DTE to loop through the projects, compile their assemblies, then reflect over those assemblies to get their referenced assemblies and cross-examine the .csproj file. Problem: since the .dll/.exe I loaded up with Reflection doesn't unload until the app domian unloads, it is now locked and the projects can't be built again because VS tries to re-create the files (all standard stuff). I have tried creating temporary files, then reflecting over them...no worky, still have locked original files (I totally don’t understand that BTW). Now I am now going down the path of creating a temporary AppDomain to load the files into and then destroy. I am having problems loading the files though: The way I understand AddDomain.Load is that I should create and send a byte array of the assembly to it. I do that: FileStream fs = new FileStream(assemblyFile, FileMode.Open); byte[] assemblyFileBuffer = new byte[(int)fs.Length]; fs.Read(assemblyFileBuffer, 0, assemblyFileBuffer.Length); fs.Close(); AppDomainSetup domainSetup = new AppDomainSetup(); domainSetup.ApplicationBase = assemblyFileInfo.Directory.FullName; AppDomain tempAppDomain = AppDomain.CreateDomain("TempAppDomain", null, domainSetup); Assembly projectAssembly = tempAppDomain.Load(assemblyFileBuffer); The last line throws an exception: "Could not load file or assembly 'WindowsFormsApplication1, Version=1.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies. The system cannot find the file specified.":"WindowsFormsApplication3, Version=1.0.0.0, Culture=neutral, PublicKeyToken=null"}" Any help or thoughts would be greatly appreciated. My head is lopsided from beating it against the wall... Thanks, Dan

  Read the article

• Trouble with building up a string in Clojure

  - by Aki Iskandar

  Hi gang - [this may seem like my problem is with Compojure, but it isn't - it's with Clojure] I've been pulling my hair out on this seemingly simple issue - but am getting nowhere. I am playing with Compojure (a light web framework for Clojure) and I would just like to generate a web page showing showing my list of todos that are in a PostgreSQL database. The code snippets are below (left out the database connection, query, etc - but that part isn't needed because specific issue is that the resulting HTML shows nothing between the <body> and </body> tags). As a test, I tried hard-coding the string in the call to main-layout, like this: (html (main-layout "Aki's Todos" "Haircut<br>Study Clojure<br>Answer a question on Stackoverfolw")) - and it works fine. So the real issue is that I do not believe I know how to build up a string in Clojure. Not the idiomatic way, and not by calling out to Java's StringBuilder either - as I have attempted to do in the code below. A virtual beer, and a big upvote to whoever can solve it! Many thanks! ============================================================= ;The master template (a very simple POC for now, but can expand on it later) (defn main-layout "This is one of the html layouts for the pages assets - just like a master page" [title body] (html [:html [:head [:title title] (include-js "todos.js") (include-css "todos.css")] [:body body]])) (defn show-all-todos "This function will generate the todos HTML table and call the layout function" [] (let [rs (select-all-todos) sbHTML (new StringBuilder)] (for [rec rs] (.append sbHTML (str rec "<br><br>"))) (html (main-layout "Aki's Todos" (.toString sbHTML))))) ============================================================= Again, the result is a web page but with nothing between the body tags. If I replace the code in the for loop with println statements, and direct the code to the repl - forgetting about the web page stuff (ie. the call to main-layout), the resultset gets printed - BUT - the issue is with building up the string. Thanks again. ~Aki

  Read the article

• XCode enum woes

  - by Raconteur

  Hi gang, I thought I had this sorted, but I am still missing something. Very simply, I have a Settings class that hold a DAO (sitting on a plist). I want to have a couple of enums for the settings for convenience and readability, such as GamePlayType and DifficultyLevel. Right now I am defining them in the Settings.h file above the @interface line as such: typedef enum { EASY, NORMAL, HARD } DifficultyLevel; and typedef enum { SET_NUMBER_OF_MOVES, TO_COMPLETION } GamePlayType; If I access them from within the Settings class like: - (int)gridSizeForLOD { switch ([self difficultyLevel]) { case EASY: return GRID_SIZE_EASY; case NORMAL: return GRID_SIZE_NORMAL; case HARD: return GRID_SIZE_HARD; default: return GRID_SIZE_NORMAL; } } everything is fine. But, if I try to access them outside of the Settings class, let's say in my main view controller class, like this: if (([settings gameType] == SET_NUMBER_OF_MOVES) && (numMoves == [settings numMovesForLOD])) { [self showLoseScreen]; } I get errors (like EXC_BAD_ACCESS) or the condition always fails. Am I doing something incorrectly? Also, I should point out that I have this code for the call to gameType (which lives in the Settings class): - (GamePlayType)gameType { return [dao gameType]; } and the DAO implements gameType like this: - (int)gameType { return (settingsContent != nil) ? [[settingsContent objectForKey:@"Game Type"] intValue] : 0; } I know I have the DAO returning an int instead of a GamePlayType, but A) the problem I am describing arose there when I tried to use the "proper" data type, and B) I did not think it would matter since the enum is just a bunch of named ints, right? Any help, greatly appreciated. I really want to understand this thoroughly, and something is eluding me... Cheers, Chris

  Read the article

• Parallelism in .NET – Part 5, Partitioning of Work

  - by Reed

  When parallelizing any routine, we start by decomposing the problem.  Once the problem is understood, we need to break our work into separate tasks, so each task can be run on a different processing element.  This process is called partitioning. Partitioning our tasks is a challenging feat.  There are opposing forces at work here: too many partitions adds overhead, too few partitions leaves processors idle.  Trying to work the perfect balance between the two extremes is the goal for which we should aim.  Luckily, the Task Parallel Library automatically handles much of this process.  However, there are situations where the default partitioning may not be appropriate, and knowledge of our routines may allow us to guide the framework to making better decisions. First off, I’d like to say that this is a more advanced topic.  It is perfectly acceptable to use the parallel constructs in the framework without considering the partitioning taking place.  The default behavior in the Task Parallel Library is very well-behaved, even for unusual work loads, and should rarely be adjusted.  I have found few situations where the default partitioning behavior in the TPL is not as good or better than my own hand-written partitioning routines, and recommend using the defaults unless there is a strong, measured, and profiled reason to avoid using them.  However, understanding partitioning, and how the TPL partitions your data, helps in understanding the proper usage of the TPL. I indirectly mentioned partitioning while discussing aggregation.  Typically, our systems will have a limited number of Processing Elements (PE), which is the terminology used for hardware capable of processing a stream of instructions.  For example, in a standard Intel i7 system, there are four processor cores, each of which has two potential hardware threads due to Hyperthreading.  This gives us a total of 8 PEs – theoretically, we can have up to eight operations occurring concurrently within our system. In order to fully exploit this power, we need to partition our work into Tasks.  A task is a simple set of instructions that can be run on a PE.  Ideally, we want to have at least one task per PE in the system, since fewer tasks means that some of our processing power will be sitting idle.  A naive implementation would be to just take our data, and partition it with one element in our collection being treated as one task.  When we loop through our collection in parallel, using this approach, we’d just process one item at a time, then reuse that thread to process the next, etc.  There’s a flaw in this approach, however.  It will tend to be slower than necessary, often slower than processing the data serially. The problem is that there is overhead associated with each task.  When we take a simple foreach loop body and implement it using the TPL, we add overhead.  First, we change the body from a simple statement to a delegate, which must be invoked.  In order to invoke the delegate on a separate thread, the delegate gets added to the ThreadPool’s current work queue, and the ThreadPool must pull this off the queue, assign it to a free thread, then execute it.  If our collection had one million elements, the overhead of trying to spawn one million tasks would destroy our performance. The answer, here, is to partition our collection into groups, and have each group of elements treated as a single task.  By adding a partitioning step, we can break our total work into small enough tasks to keep our processors busy, but large enough tasks to avoid overburdening the ThreadPool.  There are two clear, opposing goals here: Always try to keep each processor working, but also try to keep the individual partitions as large as possible. When using Parallel.For, the partitioning is always handled automatically.  At first, partitioning here seems simple.  A naive implementation would merely split the total element count up by the number of PEs in the system, and assign a chunk of data to each processor.  Many hand-written partitioning schemes work in this exactly manner.  This perfectly balanced, static partitioning scheme works very well if the amount of work is constant for each element.  However, this is rarely the case.  Often, the length of time required to process an element grows as we progress through the collection, especially if we’re doing numerical computations.  In this case, the first PEs will finish early, and sit idle waiting on the last chunks to finish.  Sometimes, work can decrease as we progress, since previous computations may be used to speed up later computations.  In this situation, the first chunks will be working far longer than the last chunks.  In order to balance the workload, many implementations create many small chunks, and reuse threads.  This adds overhead, but does provide better load balancing, which in turn improves performance. The Task Parallel Library handles this more elaborately.  Chunks are determined at runtime, and start small.  They grow slowly over time, getting larger and larger.  This tends to lead to a near optimum load balancing, even in odd cases such as increasing or decreasing workloads.  Parallel.ForEach is a bit more complicated, however. When working with a generic IEnumerable<T>, the number of items required for processing is not known in advance, and must be discovered at runtime.  In addition, since we don’t have direct access to each element, the scheduler must enumerate the collection to process it.  Since IEnumerable<T> is not thread safe, it must lock on elements as it enumerates, create temporary collections for each chunk to process, and schedule this out.  By default, it uses a partitioning method similar to the one described above.  We can see this directly by looking at the Visual Partitioning sample shipped by the Task Parallel Library team, and available as part of the Samples for Parallel Programming.  When we run the sample, with four cores and the default, Load Balancing partitioning scheme, we see this: The colored bands represent each processing core.  You can see that, when we started (at the top), we begin with very small bands of color.  As the routine progresses through the Parallel.ForEach, the chunks get larger and larger (seen by larger and larger stripes). Most of the time, this is fantastic behavior, and most likely will out perform any custom written partitioning.  However, if your routine is not scaling well, it may be due to a failure in the default partitioning to handle your specific case.  With prior knowledge about your work, it may be possible to partition data more meaningfully than the default Partitioner. There is the option to use an overload of Parallel.ForEach which takes a Partitioner<T> instance.  The Partitioner<T> class is an abstract class which allows for both static and dynamic partitioning.  By overriding Partitioner<T>.SupportsDynamicPartitions, you can specify whether a dynamic approach is available.  If not, your custom Partitioner<T> subclass would override GetPartitions(int), which returns a list of IEnumerator<T> instances.  These are then used by the Parallel class to split work up amongst processors.  When dynamic partitioning is available, GetDynamicPartitions() is used, which returns an IEnumerable<T> for each partition.  If you do decide to implement your own Partitioner<T>, keep in mind the goals and tradeoffs of different partitioning strategies, and design appropriately. The Samples for Parallel Programming project includes a ChunkPartitioner class in the ParallelExtensionsExtras project.  This provides example code for implementing your own, custom allocation strategies, including a static allocator of a given chunk size.  Although implementing your own Partitioner<T> is possible, as I mentioned above, this is rarely required or useful in practice.  The default behavior of the TPL is very good, often better than any hand written partitioning strategy.

  Read the article

• Analytic functions – they’re not aggregates

  - by Rob Farley

  SQL 2012 brings us a bunch of new analytic functions, together with enhancements to the OVER clause. People who have known me over the years will remember that I’m a big fan of the OVER clause and the types of things that it brings us when applied to aggregate functions, as well as the ranking functions that it enables. The OVER clause was introduced in SQL Server 2005, and remained frustratingly unchanged until SQL Server 2012. This post is going to look at a particular aspect of the analytic functions though (not the enhancements to the OVER clause). When I give presentations about the analytic functions around Australia as part of the tour of SQL Saturdays (starting in Brisbane this Thursday), and in Chicago next month, I’ll make sure it’s sufficiently well described. But for this post – I’m going to skip that and assume you get it. The analytic functions introduced in SQL 2012 seem to come in pairs – FIRST_VALUE and LAST_VALUE, LAG and LEAD, CUME_DIST and PERCENT_RANK, PERCENTILE_CONT and PERCENTILE_DISC. Perhaps frustratingly, they take slightly different forms as well. The ones I want to look at now are FIRST_VALUE and LAST_VALUE, and PERCENTILE_CONT and PERCENTILE_DISC. The reason I’m pulling this ones out is that they always produce the same result within their partitions (if you’re applying them to the whole partition). Consider the following query: SELECT     YEAR(OrderDate),     FIRST_VALUE(TotalDue)         OVER (PARTITION BY YEAR(OrderDate)               ORDER BY OrderDate, SalesOrderID               RANGE BETWEEN UNBOUNDED PRECEDING                         AND UNBOUNDED FOLLOWING),     LAST_VALUE(TotalDue)         OVER (PARTITION BY YEAR(OrderDate)               ORDER BY OrderDate, SalesOrderID               RANGE BETWEEN UNBOUNDED PRECEDING                         AND UNBOUNDED FOLLOWING),     PERCENTILE_CONT(0.95)         WITHIN GROUP (ORDER BY TotalDue)         OVER (PARTITION BY YEAR(OrderDate)),     PERCENTILE_DISC(0.95)         WITHIN GROUP (ORDER BY TotalDue)         OVER (PARTITION BY YEAR(OrderDate)) FROM Sales.SalesOrderHeader ; This is designed to get the TotalDue for the first order of the year, the last order of the year, and also the 95% percentile, using both the continuous and discrete methods (‘discrete’ means it picks the closest one from the values available – ‘continuous’ means it will happily use something between, similar to what you would do for a traditional median of four values). I’m sure you can imagine the results – a different value for each field, but within each year, all the rows the same. Notice that I’m not grouping by the year. Nor am I filtering. This query gives us a result for every row in the SalesOrderHeader table – 31465 in this case (using the original AdventureWorks that dates back to the SQL 2005 days). The RANGE BETWEEN bit in FIRST_VALUE and LAST_VALUE is needed to make sure that we’re considering all the rows available. If we don’t specify that, it assumes we only mean “RANGE BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW”, which means that LAST_VALUE ends up being the row we’re looking at. At this point you might think about other environments such as Access or Reporting Services, and remember aggregate functions like FIRST. We really should be able to do something like: SELECT     YEAR(OrderDate),     FIRST_VALUE(TotalDue)         OVER (PARTITION BY YEAR(OrderDate)               ORDER BY OrderDate, SalesOrderID               RANGE BETWEEN UNBOUNDED PRECEDING                         AND UNBOUNDED FOLLOWING) FROM Sales.SalesOrderHeader GROUP BY YEAR(OrderDate) ; But you can’t. You get that age-old error: Msg 8120, Level 16, State 1, Line 5 Column 'Sales.SalesOrderHeader.OrderDate' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause. Msg 8120, Level 16, State 1, Line 5 Column 'Sales.SalesOrderHeader.SalesOrderID' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause. Hmm. You see, FIRST_VALUE isn’t an aggregate function. None of these analytic functions are. There are too many things involved for SQL to realise that the values produced might be identical within the group. Furthermore, you can’t even surround it in a MAX. Then you get a different error, telling you that you can’t use windowed functions in the context of an aggregate. And so we end up grouping by doing a DISTINCT. SELECT DISTINCT     YEAR(OrderDate),         FIRST_VALUE(TotalDue)              OVER (PARTITION BY YEAR(OrderDate)                   ORDER BY OrderDate, SalesOrderID                   RANGE BETWEEN UNBOUNDED PRECEDING                             AND UNBOUNDED FOLLOWING),         LAST_VALUE(TotalDue)             OVER (PARTITION BY YEAR(OrderDate)                   ORDER BY OrderDate, SalesOrderID                   RANGE BETWEEN UNBOUNDED PRECEDING                             AND UNBOUNDED FOLLOWING),     PERCENTILE_CONT(0.95)          WITHIN GROUP (ORDER BY TotalDue)         OVER (PARTITION BY YEAR(OrderDate)),     PERCENTILE_DISC(0.95)         WITHIN GROUP (ORDER BY TotalDue)         OVER (PARTITION BY YEAR(OrderDate)) FROM Sales.SalesOrderHeader ; I’m sorry. It’s just the way it goes. Hopefully it’ll change the future, but for now, it’s what you’ll have to do. If we look in the execution plan, we see that it’s incredibly ugly, and actually works out the results of these analytic functions for all 31465 rows, finally performing the distinct operation to convert it into the four rows we get in the results. You might be able to achieve a better plan using things like TOP, or the kind of calculation that I used in http://sqlblog.com/blogs/rob_farley/archive/2011/08/23/t-sql-thoughts-about-the-95th-percentile.aspx (which is how PERCENTILE_CONT works), but it’s definitely convenient to use these functions, and in time, I’m sure we’ll see good improvements in the way that they are implemented. Oh, and this post should be good for fellow SQL Server MVP Nigel Sammy’s T-SQL Tuesday this month.

  Read the article

• SQLAuthority News – Pluralsight Course Review – Practices for Software Startups – Part 2 of 2

  - by pinaldave

  This is the second part of the two part series of Practices for Software Startup Pluralsight Course. Please read the first part of this series over here. The course is written by Stephen Forte (Blog | Twitter). Stephen Forte is the Chief Strategy Officer of the venture backed company, Telerik. Personal Learning Schedule After these three sessions it was 6:30 am and time to do my own blog.  But for the rest of the day, I kept thinking about the course, and wanted to go back and finish.  I was wishing that I had woken up at 3 am so I could finish all at one go.  All day long I was digesting what I had learned.  At 10 pm, after my daughter had gone to bed, I sighed on again.  I was not disappointed by the long wait.  As I mentioned before, Stephen has started four to six companies, and all of them are very successful today. Here is the video I promised yesterday – it discusses the importance of Right Sizing Your Startup. The Heartbeat of Startup – Technology Stephen has combined all technology knowledge into one 30 minute session.  He discussed  how to start your project, how to deal with opinions, and how to deal with multiple ideas – every start up has multiple directions it can go. He spent a lot of time emphasized deciding which direction to go and how to decide which will be the best for you.  He called it a continuous development cycle. One of the biggest hazards for a start-up company is one person deciding the direction the company will go, until down the road another team member announces that there is a glitch in their part of the work and that everyone will have to start over.  Even though a team of two or five people can move quickly, often the decision has gone too long and cannot be easily fixed.   Stephen used an example from his own life:  he was biased for one type of technology, and his teammate for another.  In the end they opted for his teammate’s  choice , and in the end it was a good decision, even though he was unfamiliar with that particular program.  He argues that technology should not be a barrier to progress, that you cannot rely on your experience only.  This really spoke to me because I am a big fan of SQL, but I know there is more out there, and I should be more open to it.  I give my thanks to Stephen, I learned something in this module besides startups. Money, Success and Epic Win! The longest, but most interesting, the module was funding your start-up.  You need to fund the start-up right at the very beginning, if not done right you will run into trouble.  The good news is that a few years ago start-ups required a lot more money – think millions of dollars – but now start-ups can get off the ground for thousands.  Stephen used an example of a company that years ago would have needed a million dollars, but today could be started for $600.  It is true that things have changed, but you still need money.  For $600 you can start small and add dynamically, as needed.  But the truth is that if you have $600, $6000, or $6 million, it will be spent.  Don’t think of it as trying to save money, think of it as investing in your future.   You will need money, and you will need to (quickly) decide what you do with the money: shares, stakeholders, investing in a team, hiring a CEO.  This is so important because once you have money and start the company, the company IS your money.  It is your biggest currency – having a percentage of ownership in the company.  Investors will want percentages as repayment for their investment, and they will want a say in the business as well.  You will have to decide how far you will dilute your shares, and how the company will be divided, if at all.  If you don’t plan in advance, you will find that after gaining three or four investors, suddenly you are the minority owner in your own dream.  You need to understand funding carefully.  This single module is worth all the money you would have spent on the whole course alone.  I encourage everyone to listen to this single module even if they don’t watch any of the others.     Press End to Start the Game – Exists! The final module is exit strategies.  You did all this work, dealt with all political and legal issues.  What are you going to get out of it? The answer is simple: money.  Maybe you want your company to be bought out, for you talent to bring you a profit.  You can sell the company to someone and still head it.  Many options are available.  You could sell and still work as an employee but no longer own the company.  There are many exit strategies.  This is where all your hard work comes into play.  It is important not to feel fooled at any step.  There are so many good ideas that end up in the garbage because of poor planning, so that if you find yourself successful, you don’t want to blow it at this step!  The exit is important.  I thought that this aspect of the course was completely unique, and I loved Stephen’s point of view.  I was lost deep in thought after this module ended.  I actually took two hours worth of notes on this section alone – and it was only a three hour course.  I am planning on attending this course one more time next week, just to catch up on all the small bits of wisdom I’m sure I missed. Thank you Stephen for bringing your real world experience with us!  I recommend that everyone attends this course, even if they don’t want to begin their own start-up company. It was indeed a long day for me. Do not forget to read part 1 of this story and attend course Practices for Software Startup Pluralsight Course. Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: Best Practices, PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

  Read the article

• Using Unity – Part 6

  - by nmarun

  This is the last of the ‘Unity’ series and I’ll be talking about generics here. If you’ve been following the previous articles, you must have noticed that I’m just adding more and more ‘Product’ classes to the project. I’ll change that trend in this blog where I’ll be adding an ICaller interface and a Caller class. 1: public interface ICaller<T> where T : IProduct 2: { 3: string CallMethod<T>(string typeName); 4: } 5:  6: public class Caller<T> : ICaller<T> where T:IProduct 7: { 8: public string CallMethod<T>(string typeName) 9: { 10: //... 11: } 12: } We’ll fill-in the implementation of the CallMethod in a few, but first, here’s what we’re going to do: create an instance of the Caller class pass it the IProduct as a generic parameter in the CallMethod method, we’ll use Unity to dynamically create an instance of IProduct implemented object I need to add the config information for ICaller and Caller types. 1: <typeAlias alias="ICaller`1" type="ProductModel.ICaller`1, ProductModel" /> 2: <typeAlias alias="Caller`1" type="ProductModel.Caller`1, ProductModel" /> The .NET Framework’s convention to express generic types is ICaller`1, where the digit following the "`" matches the number of types contained in the generic type. So a generic type that contains 4 types contained in the generic type would be declared as: 1: <typeAlias alias="Caller`4" type="ProductModel.Caller`4, ProductModel" /> On my .aspx page, I have the following UI design: 1: <asp:RadioButton ID="LegacyProduct" Text="Product" runat="server" GroupName="ProductWeb" 2: AutoPostBack="true" OnCheckedChanged="RadioButton_CheckedChanged" /> 3: <br /> 4: <asp:RadioButton ID="NewProduct" Text="Product 2" runat="server" GroupName="ProductWeb" 5: AutoPostBack="true" OnCheckedChanged="RadioButton_CheckedChanged" /> 6: <br /> 7: <asp:RadioButton ID="ComplexProduct" Text="Product 3" runat="server" GroupName="ProductWeb" 8: AutoPostBack="true" OnCheckedChanged="RadioButton_CheckedChanged" /> 9: <br /> 10: <asp:RadioButton ID="ArrayConstructor" Text="Product 4" runat="server" GroupName="ProductWeb" 11: AutoPostBack="true" OnCheckedChanged="RadioButton_CheckedChanged" /> Things to note here are that all these radio buttons belong to the same GroupName => only one of these four can be clicked. Next, all four controls postback to the same ‘OnCheckedChanged’ event and lastly the ID’s point to named types of IProduct (already added to the web.config file). 1: <type type="IProduct" mapTo="Product" name="LegacyProduct" /> 2:  3: <type type="IProduct" mapTo="Product2" name="NewProduct" /> 4:  5: <type type="IProduct" mapTo="Product3" name="ComplexProduct"> 6: ... 7: </type> 8:  9: <type type="IProduct" mapTo="Product4" name="ArrayConstructor"> 10: ... 11: </type> In my calling code, I see which radio button was clicked, pass that as an argument to the CallMethod method. 1: protected void RadioButton_CheckedChanged(object sender, EventArgs e) 2: { 3: string typeName = ((RadioButton)sender).ID; 4: ICaller<IProduct> caller = unityContainer.Resolve<ICaller<IProduct>>(); 5: productDetailsLabel.Text = caller.CallMethod<IProduct>(typeName); 6: } What’s basically happening here is that the ID of the control gets passed on to the typeName which will be one of “LegacyProduct”, “NewProduct”, “ComplexProduct” or “ArrayConstructor”. I then create an instance of an ICaller and pass the typeName to it. Now, we’ll fill in the blank for the CallMethod method (sorry for the naming guys). 1: public string CallMethod<T>(string typeName) 2: { 3: IUnityContainer unityContainer = HttpContext.Current.Application["UnityContainer"] as IUnityContainer; 4: T productInstance = unityContainer.Resolve<T>(typeName); 5: return ((IProduct)productInstance).WriteProductDetails(); 6: } This is where I’ll resolve the IProduct by passing the type name and calling the WriteProductDetails() method. With all things in place, when I run the application and choose different radio buttons, the output should look something like below:          Basically this is how generics come to play in Unity. Please see the code I’ve used for this here. This marks the end of the ‘Unity’ series. I’ll definitely post any updates that I find, but for now I don’t have anything planned.

  Read the article

• PeopleSoft Upgrades, Fusion, & BI for Leading European PeopleSoft Applications Customers

  - by Mark Rosenberg

  With so many industry-leading services firms around the globe managing their businesses with PeopleSoft, it’s always an adventure setting up times and meetings for us to keep in touch with them, especially those outside of North America who often do not get to join us at Oracle OpenWorld. Fortunately, during the first two weeks of May, Nigel Woodland (Oracle’s Service Industries Director for the EMEA region) and I successfully blocked off our calendars to visit seven different customers spanning four countries in Western Europe. We met executives and leaders at four Staffing industry firms, two Professional Services firms that engage in consulting and auditing, and a Financial Services firm. As we shared the latest information regarding product capabilities and plans, we also gained valuable insight into the hot technology topics facing these businesses. What we heard was both informative and inspiring, and I suspect other Oracle PeopleSoft applications customers can benefit from one or more of the following observations from our trip. Great IT Plans Get Executed When You Respect the Users Each of our visits followed roughly the same pattern. After introductions, Nigel outlined Oracle’s product and technology strategy, including a discussion of how we at Oracle invest in each layer of the “technology stack” to provide customers with unprecedented business management capabilities and choice. Then, I provided the specifics of the PeopleSoft product line’s investment strategy, detailing the dramatic number of rich usability and functionality enhancements added to release 9.1 since its general availability in 2009 and the game-changing capabilities slated for 9.2. What was most exciting about each of these discussions was that shortly after my talking about what customers can do with release 9.1 right now to drive up user productivity and satisfaction, I saw the wheels turning in the minds of our audiences. Business analyst and end user-configurable tools and technologies, such as WorkCenters and the Related Action Framework, that provide the ability to tailor a “central command center” to the exact needs of each recruiter, biller, and every other role in the organization were exactly what each of our customers had been looking for. Every one of our audiences agreed that these tools which demonstrate a respect for the user would finally help IT pole vault over the wall of resistance that users had often raised in the past. With these new user-focused capabilities, IT is positioned to definitively partner with the business, instead of drag the business along, to unlock the value of their investment in PeopleSoft. This topic of respecting the user emerged during our very first visit, which was at Vital Services Group at their Head Office “The Mill” in Manchester, England. (If you are a student of architecture and are ever in Manchester, you should stop in to see this amazingly renovated old mill building.) I had just finished explaining our PeopleSoft 9.2 roadmap, and Mike Code, PeopleSoft Systems Manager for this innovative staffing company, said, “Mark, the new features you’ve shown us in 9.1/9.2 are very relevant to our business. As we forge ahead with the 9.1 upgrade, the ability to configure a targeted user interface with WorkCenters, Related Actions, Pivot Grids, and Alerts will enable us to satisfy the business that this upgrade is for them and will deliver tangible benefits. In fact, you’ve highlighted that we need to start talking to the business to keep up the momentum to start reviewing the 9.2 upgrade after we get to 9.1, because as much as 9.1 and PeopleTools 8.52 offers, what you’ve shown us for 9.2 is what we’ve envisioned was ultimately possible with our investment in PeopleSoft applications.” We also received valuable feedback about our investment for the Staffing industry when we visited with Hans Wanders, CIO of Randstad (the second largest Staffing company in the world) in the Netherlands. After our visit, Hans noted, “It was very interesting to see how the PeopleSoft applications have developed. I was truly impressed by many of the new developments.” Hans and Mike, sincere thanks for the validation that our team’s hard work and dedication to “respecting the users” is worth the effort! Co-existence of PeopleSoft and Fusion Applications Just Makes Sense As a “product person,” one of the most rewarding things about visiting customers is that they actually want to talk to me. Sometimes, they want to discuss a product area that we need to enhance; other times, they are interested in learning how to extract more value from their applications; and still others, they want to tell me how they are using the applications to drive real value for the business. During this trip, I was very pleased to hear that several of our customers not only thought the co-existence of Fusion applications alongside PeopleSoft applications made sense in theory, but also that they were aggressively looking at how to deploy one or more Fusion applications alongside their PeopleSoft HCM and FSCM applications. The most common deployment plan in the works by three of the organizations is to upgrade to PeopleSoft 9.1 or 9.2, and then adopt one of the new Fusion HCM applications, such as Fusion Performance Management or the full suite of  Fusion Talent Management. For example, during an applications upgrade planning discussion with the staffing company Hays plc., Mark Thomas, who is Hays’ UK IT Director, commented, “We are very excited about where we can go with the latest versions of the PeopleSoft applications in conjunction with Fusion Talent Management.” Needless to say, this news was very encouraging, because it reiterated that our applications investment strategy makes good business sense for our customers. Next Generation Business Intelligence Is the Key to the Future The third, and perhaps most exciting, lesson I learned during this journey is that our audiences already know that the latest generation of Business Intelligence technologies will be the “secret sauce” for organizations to transform business in radical ways. While a number of the organizations we visited on the trip have deployed or are deploying Oracle Business Intelligence Enterprise Edition and the associated analytics applications to provide dashboards of easy-to-understand, user-configurable metrics that help optimize business performance according to current operating procedures, what’s most exciting to them is being able to use Business Intelligence to change the way an organization does business, grows revenue, and makes a profit. In particular, several executives we met asked whether we can help them minimize the need to have perfectly structured data and at the same time generate analytics that improve order fulfillment decision-making. To them, the path to future growth lies in having the ability to analyze unstructured data rapidly and intuitively and leveraging technology’s ability to detect patterns that a human cannot reasonably be expected to see. For illustrative purposes, here is a good example of a business problem where analyzing a combination of structured and unstructured data can produce better results. If you have a resource manager trying to decide which person would be the best fit for an assignment in terms of ensuring (a) client satisfaction, (b) the individual’s satisfaction with the work, (c) least travel distance, and (d) highest margin, you traditionally compare resource qualifications to assignment needs, calculate margins on past work with the client, and measure distances. To perform these comparisons, you are likely to need the organization to have profiles setup, people ranked against profiles, margin targets setup, margins measured, distances setup, distances measured, and more. As you can imagine, this requires organizations to plan and implement data setup, capture, and quality management initiatives to ensure that dependable information is available to support resourcing analysis and decisions. In the fast-paced, tight-budget world in which most organizations operate today, the effort and discipline required to maintain high-quality, structured data like those described in the above example are certainly not desirable and in some cases are not feasible. You can imagine how intrigued our audiences were when I informed them that we are ready to help them analyze volumes of unstructured data, detect trends, and produce recommendations. Our discussions delved into examples of how the firms could leverage Oracle’s Secure Enterprise Search and Endeca technologies to keyword search against, compare, and learn from unstructured resource and assignment data. We also considered examples of how they could employ Oracle Real-Time Decisions to generate statistically significant recommendations based on similar resourcing scenarios that have produced the desired satisfaction and profit margin results. --- Although I had almost no time for sight-seeing during this trip to Europe, I have to say that it may have been one of the most energizing and engaging trips of my career. Showing these dedicated customers how they can give every user a uniquely tailored set of tools and address business problems in ways that have to date been impossible made the journey across the Atlantic more than worth it. If any of these three topics intrigue you, I’d recommend you contact your Oracle applications representative to arrange for more detailed discussions with the appropriate members of our organization.

  Read the article

• Master Data Management and Cloud Computing

  - by david.butler(at)oracle.com

  Cloud Computing is all the rage these days. There are many reasons why this is so. But like its predecessor, Service Oriented Architecture, it can fall on hard times if the underlying data is left unmanaged. Master Data Management is the perfect Cloud companion. It can materially increase the chances for successful Cloud initiatives. In this blog, I'll review the nature of the Cloud and show how MDM fits in.   Here's the National Institute of Standards and Technology Cloud definition: •          Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or service provider interaction.   Cloud architectures have three main layers: applications or Software as a Service (SaaS), Platforms as a Service (PaaS), and Infrastructure as a Service (IaaS). SaaS generally refers to applications that are delivered to end-users over the Internet. Oracle CRM On Demand is an example of a SaaS application. Today there are hundreds of SaaS providers covering a wide variety of applications including Salesforce.com, Workday, and Netsuite. Oracle MDM applications are located in this layer of Oracle's On Demand enterprise Cloud platform. We call it Master Data as a Service (MDaaS). PaaS generally refers to an application deployment platform delivered as a service. They are often built on a grid computing architecture and include database and middleware. Oracle Fusion Middleware is in this category and includes the SOA and Data Integration products used to connect SaaS applications including MDM. Finally, IaaS generally refers to computing hardware (servers, storage and network) delivered as a service.  This typically includes the associated software as well: operating systems, virtualization, clustering, etc.    Cloud Computing benefits are compelling for a large number of organizations. These include significant cost savings, increased flexibility, and fast deployments. Cost advantages include paying for just what you use. This is especially critical for organizations with variable or seasonal usage. Companies don't have to invest to support peak computing periods. Costs are also more predictable and controllable. Increased agility includes access to the latest technology and experts without making significant up front investments.   While Cloud Computing is certainly very alluring with a clear value proposition, it is not without its challenges. An IDC survey of 244 IT executives/CIOs and their line-of-business (LOB) colleagues identified a number of issues:   Security - 74% identified security as an issue involving data privacy and resource access control. Integration - 61% found that it is hard to integrate Cloud Apps with in-house applications. Operational Costs - 50% are worried that On Demand will actually cost more given the impact of poor data quality on the rest of the enterprise. Compliance - 49% felt that compliance with required regulatory, legal and general industry requirements (such as PCI, HIPAA and Sarbanes-Oxley) would be a major issue. When control is lost, the ability of a provider to directly manage how and where data is deployed, used and destroyed is negatively impacted.  There are others, but I singled out these four top issues because Master Data Management, properly incorporated into a Cloud Computing infrastructure, can significantly ameliorate all of these problems. Cloud Computing can literally rain raw data across the enterprise.   According to fellow blogger, Mike Ferguson, "the fracturing of data caused by the adoption of cloud computing raises the importance of MDM in keeping disparate data synchronized."   David Linthicum, CTO Blue Mountain Labs blogs that "the lack of MDM will become more of an issue as cloud computing rises. We're moving from complex federated on-premise systems, to complex federated on-premise and cloud-delivered systems."    Left unmanaged, non-standard, inconsistent, ungoverned data with questionable quality can pollute analytical systems, increase operational costs, and reduce the ROI in Cloud and On-Premise applications. As cloud computing becomes more relevant, and more data, applications, services, and processes are moved out to cloud computing platforms, the need for MDM becomes ever more important. Oracle's MDM suite is designed to deal with all four of the above Cloud issues listed in the IDC survey.   Security - MDM manages all master data attribute privacy and resource access control issues. Integration - MDM pre-integrates Cloud Apps with each other and with On Premise applications at the data level. Operational Costs - MDM significantly reduces operational costs by increasing data quality, thereby improving enterprise business processes efficiency. Compliance - MDM, with its built in Data Governance capabilities, insures that the data is governed according to organizational standards. This facilitates rapid and accurate reporting for compliance purposes. Oracle MDM creates governed high quality master data. A unified cleansed and standardized data view is produced. The Oracle Customer Hub creates a single view of the customer. The Oracle Product Hub creates high quality product data designed to support all go-to-market processes. Oracle Supplier Hub dramatically reduces the chances of 'supplier exceptions'. Oracle Site Hub masters locations. And Oracle Hyperion Data Relationship Management masters financial reference data and manages enterprise hierarchies across operational areas from ERP to EPM and CRM to SCM. Oracle Fusion Middleware connects Cloud and On Premise applications to MDM Hubs and brings high quality master data to your enterprise business processes.   An independent analyst once said "Poor data quality is like dirt on the windshield. You may be able to drive for a long time with slowly degrading vision, but at some point, you either have to stop and clear the windshield or risk everything."  Cloud Computing has the potential to significantly degrade data quality across the enterprise over time. Deploying a Master Data Management solution prior to or in conjunction with a move to the Cloud can insure that the data flowing into the enterprise from the Cloud is clean and governed. This will in turn insure that expected returns on the investment in Cloud Computing will be realized.       Oracle MDM has proven its metal in this area and has the customers to back that up. In fact, I will be hosting a webcast on Tuesday, April 10th at 10 am PT with one of our top Cloud customers, the Church Pension Group. They have moved all mainline applications to a hosted model and use Oracle MDM to insure the master data is managed and cleansed before it is propagated to other cloud and internal systems. I invite you join Martin Hossfeld, VP, IT Operations, and Danette Patterson, Enterprise Data Manager as they review business drivers for MDM and hosted applications, how they did it, the benefits achieved, and lessons learned. You can register for this free webcast here.  Hope to see you there.

  Read the article

• The blocking nature of aggregates

  - by Rob Farley

  I wrote a post recently about how query tuning isn’t just about how quickly the query runs – that if you have something (such as SSIS) that is consuming your data (and probably introducing a bottleneck), then it might be more important to have a query which focuses on getting the first bit of data out. You can read that post here.  In particular, we looked at two operators that could be used to ensure that a query returns only Distinct rows. and The Sort operator pulls in all the data, sorts it (discarding duplicates), and then pushes out the remaining rows. The Hash Match operator performs a Hashing function on each row as it comes in, and then looks to see if it’s created a Hash it’s seen before. If not, it pushes the row out. The Sort method is quicker, but has to wait until it’s gathered all the data before it can do the sort, and therefore blocks the data flow. But that was my last post. This one’s a bit different. This post is going to look at how Aggregate functions work, which ties nicely into this month’s T-SQL Tuesday. I’ve frequently explained about the fact that DISTINCT and GROUP BY are essentially the same function, although DISTINCT is the poorer cousin because you have less control over it, and you can’t apply aggregate functions. Just like the operators used for Distinct, there are different flavours of Aggregate operators – coming in blocking and non-blocking varieties. The example I like to use to explain this is a pile of playing cards. If I’m handed a pile of cards and asked to count how many cards there are in each suit, it’s going to help if the cards are already ordered. Suppose I’m playing a game of Bridge, I can easily glance at my hand and count how many there are in each suit, because I keep the pile of cards in order. Moving from left to right, I could tell you I have four Hearts in my hand, even before I’ve got to the end. By telling you that I have four Hearts as soon as I know, I demonstrate the principle of a non-blocking operation. This is known as a Stream Aggregate operation. It requires input which is sorted by whichever columns the grouping is on, and it will release a row as soon as the group changes – when I encounter a Spade, I know I don’t have any more Hearts in my hand. Alternatively, if the pile of cards are not sorted, I won’t know how many Hearts I have until I’ve looked through all the cards. In fact, to count them, I basically need to put them into little piles, and when I’ve finished making all those piles, I can count how many there are in each. Because I don’t know any of the final numbers until I’ve seen all the cards, this is blocking. This performs the aggregate function using a Hash Match. Observant readers will remember this from my Distinct example. You might remember that my earlier Hash Match operation – used for Distinct Flow – wasn’t blocking. But this one is. They’re essentially doing a similar operation, applying a Hash function to some data and seeing if the set of values have been seen before, but before, it needs more information than the mere existence of a new set of values, it needs to consider how many of them there are. A lot is dependent here on whether the data coming out of the source is sorted or not, and this is largely determined by the indexes that are being used. If you look in the Properties of an Index Scan, you’ll be able to see whether the order of the data is required by the plan. A property called Ordered will demonstrate this. In this particular example, the second plan is significantly faster, but is dependent on having ordered data. In fact, if I force a Stream Aggregate on unordered data (which I’m doing by telling it to use a different index), a Sort operation is needed, which makes my plan a lot slower. This is all very straight-forward stuff, and information that most people are fully aware of. I’m sure you’ve all read my good friend Paul White (@sql_kiwi)’s post on how the Query Optimizer chooses which type of aggregate function to apply. But let’s take a look at SQL Server Integration Services. SSIS gives us a Aggregate transformation for use in Data Flow Tasks, but it’s described as Blocking. The definitive article on Performance Tuning SSIS uses Sort and Aggregate as examples of Blocking Transformations. I’ve just shown you that Aggregate operations used by the Query Optimizer are not always blocking, but that the SSIS Aggregate component is an example of a blocking transformation. But is it always the case? After all, there are plenty of SSIS Performance Tuning talks out there that describe the value of sorted data in Data Flow Tasks, describing the IsSorted property that can be set through the Advanced Editor of your Source component. And so I set about testing the Aggregate transformation in SSIS, to prove for sure whether providing Sorted data would let the Aggregate transform behave like a Stream Aggregate. (Of course, I knew the answer already, but it helps to be able to demonstrate these things). A query that will produce a million rows in order was in order. Let me rephrase. I used a query which produced the numbers from 1 to 1000000, in a single field, ordered. The IsSorted flag was set on the source output, with the only column as SortKey 1. Performing an Aggregate function over this (counting the number of rows per distinct number) should produce an additional column with 1 in it. If this were being done in T-SQL, the ordered data would allow a Stream Aggregate to be used. In fact, if the Query Optimizer saw that the field had a Unique Index on it, it would be able to skip the Aggregate function completely, and just insert the value 1. This is a shortcut I wouldn’t be expecting from SSIS, but certainly the Stream behaviour would be nice. Unfortunately, it’s not the case. As you can see from the screenshots above, the data is pouring into the Aggregate function, and not being released until all million rows have been seen. It’s not doing a Stream Aggregate at all. This is expected behaviour. (I put that in bold, because I want you to realise this.) An SSIS transformation is a piece of code that runs. It’s a physical operation. When you write T-SQL and ask for an aggregation to be done, it’s a logical operation. The physical operation is either a Stream Aggregate or a Hash Match. In SSIS, you’re telling the system that you want a generic Aggregation, that will have to work with whatever data is passed in. I’m not saying that it wouldn’t be possible to make a sometimes-blocking aggregation component in SSIS. A Custom Component could be created which could detect whether the SortKeys columns of the input matched the Grouping columns of the Aggregation, and either call the blocking code or the non-blocking code as appropriate. One day I’ll make one of those, and publish it on my blog. I’ve done it before with a Script Component, but as Script components are single-use, I was able to handle the data knowing everything about my data flow already. As per my previous post – there are a lot of aspects in which tuning SSIS and tuning execution plans use similar concepts. In both situations, it really helps to have a feel for what’s going on behind the scenes. Considering whether an operation is blocking or not is extremely relevant to performance, and that it’s not always obvious from the surface. In a future post, I’ll show the impact of blocking v non-blocking and synchronous v asynchronous components in SSIS, using some of LobsterPot’s Script Components and Custom Components as examples. When I get that sorted, I’ll make a Stream Aggregate component available for download.

  Read the article

• The blocking nature of aggregates

  - by Rob Farley

  I wrote a post recently about how query tuning isn’t just about how quickly the query runs – that if you have something (such as SSIS) that is consuming your data (and probably introducing a bottleneck), then it might be more important to have a query which focuses on getting the first bit of data out. You can read that post here.  In particular, we looked at two operators that could be used to ensure that a query returns only Distinct rows. and The Sort operator pulls in all the data, sorts it (discarding duplicates), and then pushes out the remaining rows. The Hash Match operator performs a Hashing function on each row as it comes in, and then looks to see if it’s created a Hash it’s seen before. If not, it pushes the row out. The Sort method is quicker, but has to wait until it’s gathered all the data before it can do the sort, and therefore blocks the data flow. But that was my last post. This one’s a bit different. This post is going to look at how Aggregate functions work, which ties nicely into this month’s T-SQL Tuesday. I’ve frequently explained about the fact that DISTINCT and GROUP BY are essentially the same function, although DISTINCT is the poorer cousin because you have less control over it, and you can’t apply aggregate functions. Just like the operators used for Distinct, there are different flavours of Aggregate operators – coming in blocking and non-blocking varieties. The example I like to use to explain this is a pile of playing cards. If I’m handed a pile of cards and asked to count how many cards there are in each suit, it’s going to help if the cards are already ordered. Suppose I’m playing a game of Bridge, I can easily glance at my hand and count how many there are in each suit, because I keep the pile of cards in order. Moving from left to right, I could tell you I have four Hearts in my hand, even before I’ve got to the end. By telling you that I have four Hearts as soon as I know, I demonstrate the principle of a non-blocking operation. This is known as a Stream Aggregate operation. It requires input which is sorted by whichever columns the grouping is on, and it will release a row as soon as the group changes – when I encounter a Spade, I know I don’t have any more Hearts in my hand. Alternatively, if the pile of cards are not sorted, I won’t know how many Hearts I have until I’ve looked through all the cards. In fact, to count them, I basically need to put them into little piles, and when I’ve finished making all those piles, I can count how many there are in each. Because I don’t know any of the final numbers until I’ve seen all the cards, this is blocking. This performs the aggregate function using a Hash Match. Observant readers will remember this from my Distinct example. You might remember that my earlier Hash Match operation – used for Distinct Flow – wasn’t blocking. But this one is. They’re essentially doing a similar operation, applying a Hash function to some data and seeing if the set of values have been seen before, but before, it needs more information than the mere existence of a new set of values, it needs to consider how many of them there are. A lot is dependent here on whether the data coming out of the source is sorted or not, and this is largely determined by the indexes that are being used. If you look in the Properties of an Index Scan, you’ll be able to see whether the order of the data is required by the plan. A property called Ordered will demonstrate this. In this particular example, the second plan is significantly faster, but is dependent on having ordered data. In fact, if I force a Stream Aggregate on unordered data (which I’m doing by telling it to use a different index), a Sort operation is needed, which makes my plan a lot slower. This is all very straight-forward stuff, and information that most people are fully aware of. I’m sure you’ve all read my good friend Paul White (@sql_kiwi)’s post on how the Query Optimizer chooses which type of aggregate function to apply. But let’s take a look at SQL Server Integration Services. SSIS gives us a Aggregate transformation for use in Data Flow Tasks, but it’s described as Blocking. The definitive article on Performance Tuning SSIS uses Sort and Aggregate as examples of Blocking Transformations. I’ve just shown you that Aggregate operations used by the Query Optimizer are not always blocking, but that the SSIS Aggregate component is an example of a blocking transformation. But is it always the case? After all, there are plenty of SSIS Performance Tuning talks out there that describe the value of sorted data in Data Flow Tasks, describing the IsSorted property that can be set through the Advanced Editor of your Source component. And so I set about testing the Aggregate transformation in SSIS, to prove for sure whether providing Sorted data would let the Aggregate transform behave like a Stream Aggregate. (Of course, I knew the answer already, but it helps to be able to demonstrate these things). A query that will produce a million rows in order was in order. Let me rephrase. I used a query which produced the numbers from 1 to 1000000, in a single field, ordered. The IsSorted flag was set on the source output, with the only column as SortKey 1. Performing an Aggregate function over this (counting the number of rows per distinct number) should produce an additional column with 1 in it. If this were being done in T-SQL, the ordered data would allow a Stream Aggregate to be used. In fact, if the Query Optimizer saw that the field had a Unique Index on it, it would be able to skip the Aggregate function completely, and just insert the value 1. This is a shortcut I wouldn’t be expecting from SSIS, but certainly the Stream behaviour would be nice. Unfortunately, it’s not the case. As you can see from the screenshots above, the data is pouring into the Aggregate function, and not being released until all million rows have been seen. It’s not doing a Stream Aggregate at all. This is expected behaviour. (I put that in bold, because I want you to realise this.) An SSIS transformation is a piece of code that runs. It’s a physical operation. When you write T-SQL and ask for an aggregation to be done, it’s a logical operation. The physical operation is either a Stream Aggregate or a Hash Match. In SSIS, you’re telling the system that you want a generic Aggregation, that will have to work with whatever data is passed in. I’m not saying that it wouldn’t be possible to make a sometimes-blocking aggregation component in SSIS. A Custom Component could be created which could detect whether the SortKeys columns of the input matched the Grouping columns of the Aggregation, and either call the blocking code or the non-blocking code as appropriate. One day I’ll make one of those, and publish it on my blog. I’ve done it before with a Script Component, but as Script components are single-use, I was able to handle the data knowing everything about my data flow already. As per my previous post – there are a lot of aspects in which tuning SSIS and tuning execution plans use similar concepts. In both situations, it really helps to have a feel for what’s going on behind the scenes. Considering whether an operation is blocking or not is extremely relevant to performance, and that it’s not always obvious from the surface. In a future post, I’ll show the impact of blocking v non-blocking and synchronous v asynchronous components in SSIS, using some of LobsterPot’s Script Components and Custom Components as examples. When I get that sorted, I’ll make a Stream Aggregate component available for download.

  Read the article

• Experiences with learning Chinese

  - by Greg Low

  I've had a few friends asking me about learning Chinese and what I've found works and doesn't work. I was answering a question on a mailing list today and I thought I should post this info where it might be useful to many. The question that was initially asked was whether Rosetta Stone was useful but I've provided much more info on learning the language here. I’ve used Rosetta Stone with Chinese but it’s really hard to know whether to recommend it or not. Rosetta Stone works the same way in all languages. They show you photos and then let you both see and hear the target language and get you to work out what they’re talking about. The thinking is that that’s how children learn. However, at first, I found it very frustrating. I’d be staring at photos trying to work out what they were really trying to get at. Sometimes it’s far from obvious. I could not have survived without Google Translate open at the same time. The other weird thing is that the photos are from a mixture of countries. While that’s good in a way, it also means that they are endlessly showing pictures of something that would never happen in the target language and culture. For any language, constant interaction with a speaker of the target language is needed. Rosetta Stone has a “Studio” option. That’s the best part of the program. In my case, it lets me connect around twice a week to a live online class from Beijing. Classes usually have the teacher plus two to four students. You get some Studio access with the initial packages but need to purchase it for ongoing use. I find it very inexpensive. It seems to work out to about $70 (AUD/USD) for six months. That’s a real bargain. The other downside to Rosetta Stone is that they tend to teach very formal language, but as with other languages, that’s not how the locals speak. It might have been correct at one point but no-one actually says that. As an example, Rosetta Stone teach Gonggòng qìche (pronounced roughly like “gong gong chee chure” for bus. Most of my friends from areas like Taiwan would just say Gongche. Google Translate says Zongxiàn (pronounced somewhat like “dzong sheean”) instead. Mind you, the Rosetta Stone option isn't really as bad as "omnibus"; it's more like saying "public bus". If you say the option they provide, people would understand you. I also listen to ChinesePod in the car. They also have SpanishPod. Each podcast is about five minutes of spoken conversation. It is very good for providing current language. Another resource I use is local Meetup groups. Most cities have these and for a variety of languages. It’s way less structured (just standard conversation) but good for getting interaction. The obvious challenge for Asian languages is reading/writing. The input editors for Chinese that are part of Windows are excellent. Many of my Chinese friends speak fluently but cannot read or write. I was determined to learn to do both. For writing, I’m talking about on a computer, not with a pen. (Mind you, I can barely write English with a pen nowadays). When using Rosetta Stone, you can choose to have the Chinese words displayed in pinyin (Wo xihuan xuéxí zhongguó) or in Chinese characters (???????) or both. This year, I’ve been forcing myself to just use the Chinese characters. I use a pinyin input editor in Windows though, as it’s very fast.  (The character recognition input in the iPad is also amazing). Notice from the example that I provided above that the pronunciation of the pinyin isn’t that obvious to us at first either.  Since changing to only using characters, I find I can now read many more Chinese characters fluently. It’s a major challenge though. I can read about 300 now and yet you need around 2,500 to be able to read a newspaper fairly well. Tones are a major issue for some Asian languages. Mandarin has four tones (plus a neutral tone) and there is a major difference in meaning between two words that are spelled the same in pinyin but with different tones. For example, Ma (3rd tone?) is a horse, Ma (1st tone?) is like “mom”, and ma (neutral tone?) is a question mark and so on. Clearly you don’t want to mix these up. As in English, they also have words that do sound the same but mean different things in different contexts. What’s interesting is that even though we see two words that differ only by tone as very similar, to a native speaker, if you say the right words with the wrong tone, you might as well have said a completely different word. My wife’s dialect of Chinese has eight tones. It’s much worse. The reason I’m so keen to learn to read/write Chinese is that even though the different dialects are pronounced so differently that speakers of one dialect often cannot understand another dialect, the writing is generally the same. The only difference is that many years ago, the Chinese government created a simplified set of characters for some of the most commonly used ones. Older Chinese and most Cantonese speakers often struggle with the simplified characters. This is the simplified form of “three apples”: ????   This is the traditional form of the same words: ????  Note that two of the characters are the same but the middle two are quite different. For most languages, the best thing is to watch current movies in the target language but to watch them with the target language as subtitles, not your native language. You want to know what they actually said, not what it roughly means (which is what the English subtitle would give you). The difficulty with Asian languages like Chinese is that you have the added challenge of understanding the subtitles when they are written in the target language. I wish there were Mandarin Chinese movies with pinyin subtitles. For learning to read characters, I also recommend HanCard on the iPad. It is targeted at the HSK language proficiency levels. (I’m intending to take the first HSK exam as soon as I’m ready). Hope that info helps someone get started.  

  Read the article

• Organization &amp; Architecture UNISA Studies &ndash; Chap 4

  - by MarkPearl

  Learning Outcomes Explain the characteristics of memory systems Describe the memory hierarchy Discuss cache memory principles Discuss issues relevant to cache design Describe the cache organization of the Pentium Computer Memory Systems There are key characteristics of memory… Location – internal or external Capacity – expressed in terms of bytes Unit of Transfer – the number of bits read out of or written into memory at a time Access Method – sequential, direct, random or associative From a users perspective the two most important characteristics of memory are… Capacity Performance – access time, memory cycle time, transfer rate The trade off for memory happens along three axis… Faster access time, greater cost per bit Greater capacity, smaller cost per bit Greater capacity, slower access time This leads to people using a tiered approach in their use of memory   As one goes down the hierarchy, the following occurs… Decreasing cost per bit Increasing capacity Increasing access time Decreasing frequency of access of the memory by the processor The use of two levels of memory to reduce average access time works in principle, but only if conditions 1 to 4 apply. A variety of technologies exist that allow us to accomplish this. Thus it is possible to organize data across the hierarchy such that the percentage of accesses to each successively lower level is substantially less than that of the level above. A portion of main memory can be used as a buffer to hold data temporarily that is to be read out to disk. This is sometimes referred to as a disk cache and improves performance in two ways… Disk writes are clustered. Instead of many small transfers of data, we have a few large transfers of data. This improves disk performance and minimizes processor involvement. Some data designed for write-out may be referenced by a program before the next dump to disk. In that case the data is retrieved rapidly from the software cache rather than slowly from disk. Cache Memory Principles Cache memory is substantially faster than main memory. A caching system works as follows.. When a processor attempts to read a word of memory, a check is made to see if this in in cache memory… If it is, the data is supplied, If it is not in the cache, a block of main memory, consisting of a fixed number of words is loaded to the cache. Because of the phenomenon of locality of references, when a block of data is fetched into the cache, it is likely that there will be future references to that same memory location or to other words in the block. Elements of Cache Design While there are a large number of cache implementations, there are a few basic design elements that serve to classify and differentiate cache architectures… Cache Addresses Cache Size Mapping Function Replacement Algorithm Write Policy Line Size Number of Caches Cache Addresses Almost all non-embedded processors support virtual memory. Virtual memory in essence allows a program to address memory from a logical point of view without needing to worry about the amount of physical memory available. When virtual addresses are used the designer may choose to place the cache between the MMU (memory management unit) and the processor or between the MMU and main memory. The disadvantage of virtual memory is that most virtual memory systems supply each application with the same virtual memory address space (each application sees virtual memory starting at memory address 0), which means the cache memory must be completely flushed with each application context switch or extra bits must be added to each line of the cache to identify which virtual address space the address refers to. Cache Size We would like the size of the cache to be small enough so that the overall average cost per bit is close to that of main memory alone and large enough so that the overall average access time is close to that of the cache alone. Also, larger caches are slightly slower than smaller ones. Mapping Function Because there are fewer cache lines than main memory blocks, an algorithm is needed for mapping main memory blocks into cache lines. The choice of mapping function dictates how the cache is organized. Three techniques can be used… Direct – simplest technique, maps each block of main memory into only one possible cache line Associative – Each main memory block to be loaded into any line of the cache Set Associative – exhibits the strengths of both the direct and associative approaches while reducing their disadvantages For detailed explanations of each approach – read the text book (page 148 – 154) Replacement Algorithm For associative and set associating mapping a replacement algorithm is needed to determine which of the existing blocks in the cache must be replaced by a new block. There are four common approaches… LRU (Least recently used) FIFO (First in first out) LFU (Least frequently used) Random selection Write Policy When a block resident in the cache is to be replaced, there are two cases to consider If no writes to that block have happened in the cache – discard it If a write has occurred, a process needs to be initiated where the changes in the cache are propagated back to the main memory. There are several approaches to achieve this including… Write Through – all writes to the cache are done to the main memory as well at the point of the change Write Back – when a block is replaced, all dirty bits are written back to main memory The problem is complicated when we have multiple caches, there are techniques to accommodate for this but I have not summarized them. Line Size When a block of data is retrieved and placed in the cache, not only the desired word but also some number of adjacent words are retrieved. As the block size increases from very small to larger sizes, the hit ratio will at first increase because of the principle of locality, which states that the data in the vicinity of a referenced word are likely to be referenced in the near future. As the block size increases, more useful data are brought into cache. The hit ratio will begin to decrease as the block becomes even bigger and the probability of using the newly fetched information becomes less than the probability of using the newly fetched information that has to be replaced. Two specific effects come into play… Larger blocks reduce the number of blocks that fit into a cache. Because each block fetch overwrites older cache contents, a small number of blocks results in data being overwritten shortly after they are fetched. As a block becomes larger, each additional word is farther from the requested word and therefore less likely to be needed in the near future. The relationship between block size and hit ratio is complex, and no set approach is judged to be the best in all circumstances.   Pentium 4 and ARM cache organizations The processor core consists of four major components: Fetch/decode unit – fetches program instruction in order from the L2 cache, decodes these into a series of micro-operations, and stores the results in the L2 instruction cache Out-of-order execution logic – Schedules execution of the micro-operations subject to data dependencies and resource availability – thus micro-operations may be scheduled for execution in a different order than they were fetched from the instruction stream. As time permits, this unit schedules speculative execution of micro-operations that may be required in the future Execution units – These units execute micro-operations, fetching the required data from the L1 data cache and temporarily storing results in registers Memory subsystem – This unit includes the L2 and L3 caches and the system bus, which is used to access main memory when the L1 and L2 caches have a cache miss and to access the system I/O resources

  Read the article

• Help with Windows 7 BSOD with windbg minidump !analyze -v results

  - by Kurt Harless

  Hi gang, Windows 7 X64 Ultimate is BSODing occasionally. I suspect an overheating issue or something related to the use of my GTX-295 card that runs very hot. Here is an !analyze -v listing of the most recent minidump. Any and all help greatly appreciated. Kurt Microsoft (R) Windows Debugger Version 6.12.0002.633 AMD64 Copyright (c) Microsoft Corporation. All rights reserved. Loading Dump File [C:\Windows\Minidump\122810-31387-01.dmp] Mini Kernel Dump File: Only registers and stack trace are available Symbol search path is: SRV*c:\websymbols*http://msdl.microsoft.com/download/symbols Executable search path is: Windows 7 Kernel Version 7600 MP (8 procs) Free x64 Product: WinNt, suite: TerminalServer SingleUserTS Built by: 7600.16617.amd64fre.win7_gdr.100618-1621 Machine Name: Kernel base = 0xfffff800`03065000 PsLoadedModuleList = 0xfffff800`032a2e50 Debug session time: Tue Dec 28 11:04:03.597 2010 (UTC - 7:00) System Uptime: 2 days 2:28:40.407 Loading Kernel Symbols ............................................................... ................................................................ .............................................. Loading User Symbols Loading unloaded module list ................ ******************************************************************************* * * * Bugcheck Analysis * * * ******************************************************************************* Use !analyze -v to get detailed debugging information. BugCheck 3B, {c0000005, fffff800033b8873, fffff8800e322dc0, 0} Probably caused by : ntkrnlmp.exe ( nt!RtlCompareUnicodeStrings+c3 ) Followup: MachineOwner --------- 1: kd> !analyze -v ******************************************************************************* * * * Bugcheck Analysis * * * ******************************************************************************* SYSTEM_SERVICE_EXCEPTION (3b) An exception happened while executing a system service routine. Arguments: Arg1: 00000000c0000005, Exception code that caused the bugcheck Arg2: fffff800033b8873, Address of the instruction which caused the bugcheck Arg3: fffff8800e322dc0, Address of the context record for the exception that caused the bugcheck Arg4: 0000000000000000, zero. Debugging Details: ------------------ EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at 0x%08lx referenced memory at 0x%08lx. The memory could not be %s. FAULTING_IP: nt!RtlCompareUnicodeStrings+c3 fffff800`033b8873 488b7c2418 mov rdi,qword ptr [rsp+18h] CONTEXT: fffff8800e322dc0 -- (.cxr 0xfffff8800e322dc0) rax=0000000000000041 rbx=fffff8a015a3c1c0 rcx=0000000000000024 rdx=0000000000000003 rsi=fffff8800e3238b0 rdi=0000000000000009 rip=fffff800033b8873 rsp=fffff8800e323798 rbp=000000000000000d r8=fffff8a018cb374c r9=000000200a98fdc4 r10=fffff8800e323988 r11=fffff8800e32398e r12=fffff8a018127c18 r13=fffff8800126e550 r14=0000000000000001 r15=fffffa800abe1570 iopl=0 nv up ei pl nz ac po nc cs=0010 ss=0018 ds=002b es=002b fs=0053 gs=002b efl=00010216 nt!RtlCompareUnicodeStrings+0xc3: fffff800`033b8873 488b7c2418 mov rdi,qword ptr [rsp+18h] ss:0018:fffff880`0e3237b0=???????????????? Resetting default scope CUSTOMER_CRASH_COUNT: 1 DEFAULT_BUCKET_ID: VISTA_DRIVER_FAULT BUGCHECK_STR: 0x3B PROCESS_NAME: ccSvcHst.exe CURRENT_IRQL: 0 LAST_CONTROL_TRANSFER: from 0000000000000000 to fffff800033b8873 STACK_TEXT: fffff880`0e323798 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!RtlCompareUnicodeStrings+0xc3 FOLLOWUP_IP: nt!RtlCompareUnicodeStrings+c3 fffff800`033b8873 488b7c2418 mov rdi,qword ptr [rsp+18h] SYMBOL_STACK_INDEX: 0 SYMBOL_NAME: nt!RtlCompareUnicodeStrings+c3 FOLLOWUP_NAME: MachineOwner MODULE_NAME: nt IMAGE_NAME: ntkrnlmp.exe DEBUG_FLR_IMAGE_TIMESTAMP: 4c1c44a9 STACK_COMMAND: .cxr 0xfffff8800e322dc0 ; kb FAILURE_BUCKET_ID: X64_0x3B_nt!RtlCompareUnicodeStrings+c3 BUCKET_ID: X64_0x3B_nt!RtlCompareUnicodeStrings+c3 Followup: MachineOwner ---------

  Read the article

• Talking JavaOne with Rock Star Martijn Verburg

  - by Janice J. Heiss

  JavaOne Rock Stars, conceived in 2005, are the top-rated speakers at each JavaOne Conference. They are awarded by their peers, who, through conference surveys, recognize them for their outstanding sessions and speaking ability. Over the years many of the world’s leading Java developers have been so recognized. Martijn Verburg has, in recent years, established himself as an important mover and shaker in the Java community. His “Diabolical Developer” session at the JavaOne 2011 Conference got people’s attention by identifying some of the worst practices Java developers are prone to engage in. Among other things, he is co-leader and organizer of the thriving London Java User Group (JUG) which has more than 2,500 members, co-represents the London JUG on the Executive Committee of the Java Community Process, and leads the global effort for the Java User Group “Adopt a JSR” and “Adopt OpenJDK” programs. Career highlights include overhauling technology stacks and SDLC practices at Mizuho International, mentoring Oracle on technical community management, and running off shore development teams for AIG. He is currently CTO at jClarity, a start-up focusing on automating optimization for Java/JVM related technologies, and Product Advisor at ZeroTurnaround. He co-authored, with Ben Evans, "The Well-Grounded Java Developer" published by Manning and, as a leading authority on technical team optimization, he is in high demand at major software conferences.Verburg is participating in five sessions, a busy man indeed. Here they are: CON6152 - Modern Software Development Antipatterns (with Ben Evans) UGF10434 - JCP and OpenJDK: Using the JUGs’ “Adopt” Programs in Your Group (with Csaba Toth) BOF4047 - OpenJDK Building and Testing: Case Study—Java User Group OpenJDK Bugathon (with Ben Evans and Cecilia Borg) BOF6283 - 101 Ways to Improve Java: Why Developer Participation Matters (with Bruno Souza and Heather Vancura-Chilson) HOL6500 - Finding and Solving Java Deadlocks (with Heinz Kabutz, Kirk Pepperdine, Ellen Kraffmiller and Henri Tremblay) When I asked Verburg about the biggest mistakes Java developers tend to make, he listed three: A lack of communication -- Software development is far more a social activity than a technical one; most projects fail because of communication issues and social dynamics, not because of a bad technical decision. Sadly, many developers never learn this lesson. No source control -- Developers simply storing code in local filesystems and emailing code in order to integrate Design-driven Design -- The need for some developers to cram every design pattern from the Gang of Four (GoF) book into their source code All of which raises the question: If these practices are so bad, why do developers engage in them? “I've seen a wide gamut of reasons,” said Verburg, who lists them as: * They were never taught at high school/university that their bad habits were harmful.* They weren't mentored in their first professional roles.* They've lost passion for their craft.* They're being deliberately malicious!* They think software development is a technical activity and not a social one.* They think that they'll be able to tidy it up later.A couple of key confusions and misconceptions beset Java developers, according to Verburg. “With Java and the JVM in particular I've seen a couple of trends,” he remarked. “One is that developers think that the JVM is a magic box that will clean up their memory, make their code run fast, as well as make them cups of coffee. The JVM does help in a lot of cases, but bad code can and will still lead to terrible results! The other trend is to try and force Java (the language) to do something it's not very good at, such as rapid web development. So you get a proliferation of overly complex frameworks, libraries and techniques trying to get around the fact that Java is a monolithic, statically typed, compiled, OO environment. It's not a Golden Hammer!”I asked him about the keys to running a good Java User Group. “You need to have a ‘Why,’” he observed. “Many user groups know what they do (typically, events) and how they do it (the logistics), but what really drives users to join your group and to stay is to give them a purpose. For example, within the LJC we constantly talk about the ‘Why,’ which in our case is several whys:* Re-ignite the passion that developers have for their craft* Raise the bar of Java developers in London* We want developers to have a voice in deciding the future of Java* We want to inspire the next generation of tech leaders* To bring the disparate tech groups in London together* So we could learn from each other* We believe that the Java ecosystem forms a cornerstone of our society today -- we want to protect that for the futureLooking ahead to Java 8 Verburg expressed excitement about Lambdas. “I cannot wait for Lambdas,” he enthused. “Brian Goetz and his group are doing a great job, especially given some of the backwards compatibility that they have to maintain. It's going to remove a lot of boiler plate and yet maintain readability, plus enable massive scaling.”Check out Martijn Verburg at JavaOne if you get a chance, and, stay tuned for a longer interview yours truly did with Martijn to be publish on otn/java some time after JavaOne. Originally published on blogs.oracle.com/javaone.

  Read the article

• Talking JavaOne with Rock Star Martijn Verburg

  - by Janice J. Heiss

  JavaOne Rock Stars, conceived in 2005, are the top-rated speakers at each JavaOne Conference. They are awarded by their peers, who, through conference surveys, recognize them for their outstanding sessions and speaking ability. Over the years many of the world’s leading Java developers have been so recognized. Martijn Verburg has, in recent years, established himself as an important mover and shaker in the Java community. His “Diabolical Developer” session at the JavaOne 2011 Conference got people’s attention by identifying some of the worst practices Java developers are prone to engage in. Among other things, he is co-leader and organizer of the thriving London Java User Group (JUG) which has more than 2,500 members, co-represents the London JUG on the Executive Committee of the Java Community Process, and leads the global effort for the Java User Group “Adopt a JSR” and “Adopt OpenJDK” programs. Career highlights include overhauling technology stacks and SDLC practices at Mizuho International, mentoring Oracle on technical community management, and running off shore development teams for AIG. He is currently CTO at jClarity, a start-up focusing on automating optimization for Java/JVM related technologies, and Product Advisor at ZeroTurnaround. He co-authored, with Ben Evans, "The Well-Grounded Java Developer" published by Manning and, as a leading authority on technical team optimization, he is in high demand at major software conferences.Verburg is participating in five sessions, a busy man indeed. Here they are: CON6152 - Modern Software Development Antipatterns (with Ben Evans) UGF10434 - JCP and OpenJDK: Using the JUGs’ “Adopt” Programs in Your Group (with Csaba Toth) BOF4047 - OpenJDK Building and Testing: Case Study—Java User Group OpenJDK Bugathon (with Ben Evans and Cecilia Borg) BOF6283 - 101 Ways to Improve Java: Why Developer Participation Matters (with Bruno Souza and Heather Vancura-Chilson) HOL6500 - Finding and Solving Java Deadlocks (with Heinz Kabutz, Kirk Pepperdine, Ellen Kraffmiller and Henri Tremblay) When I asked Verburg about the biggest mistakes Java developers tend to make, he listed three: A lack of communication -- Software development is far more a social activity than a technical one; most projects fail because of communication issues and social dynamics, not because of a bad technical decision. Sadly, many developers never learn this lesson. No source control -- Developers simply storing code in local filesystems and emailing code in order to integrate Design-driven Design -- The need for some developers to cram every design pattern from the Gang of Four (GoF) book into their source code All of which raises the question: If these practices are so bad, why do developers engage in them? “I've seen a wide gamut of reasons,” said Verburg, who lists them as: * They were never taught at high school/university that their bad habits were harmful.* They weren't mentored in their first professional roles.* They've lost passion for their craft.* They're being deliberately malicious!* They think software development is a technical activity and not a social one.* They think that they'll be able to tidy it up later.A couple of key confusions and misconceptions beset Java developers, according to Verburg. “With Java and the JVM in particular I've seen a couple of trends,” he remarked. “One is that developers think that the JVM is a magic box that will clean up their memory, make their code run fast, as well as make them cups of coffee. The JVM does help in a lot of cases, but bad code can and will still lead to terrible results! The other trend is to try and force Java (the language) to do something it's not very good at, such as rapid web development. So you get a proliferation of overly complex frameworks, libraries and techniques trying to get around the fact that Java is a monolithic, statically typed, compiled, OO environment. It's not a Golden Hammer!”I asked him about the keys to running a good Java User Group. “You need to have a ‘Why,’” he observed. “Many user groups know what they do (typically, events) and how they do it (the logistics), but what really drives users to join your group and to stay is to give them a purpose. For example, within the LJC we constantly talk about the ‘Why,’ which in our case is several whys:* Re-ignite the passion that developers have for their craft* Raise the bar of Java developers in London* We want developers to have a voice in deciding the future of Java* We want to inspire the next generation of tech leaders* To bring the disparate tech groups in London together* So we could learn from each other* We believe that the Java ecosystem forms a cornerstone of our society today -- we want to protect that for the futureLooking ahead to Java 8 Verburg expressed excitement about Lambdas. “I cannot wait for Lambdas,” he enthused. “Brian Goetz and his group are doing a great job, especially given some of the backwards compatibility that they have to maintain. It's going to remove a lot of boiler plate and yet maintain readability, plus enable massive scaling.”Check out Martijn Verburg at JavaOne if you get a chance, and, stay tuned for a longer interview yours truly did with Martijn to be publish on otn/java some time after JavaOne.

  Read the article

• Becoming A Great Developer

  - by Lee Brandt

  Image via Wikipedia I’ve been doing the whole programming thing for awhile and reading and watching some of the best in the business. I have come to notice that the really great developers do a few things that (I think) makes them great. Now don’t get me wrong, I am not saying that I am one of these few. I still struggle with doing some of the things that makes one great at development. Coincidently, many of these things also make you a better person period. Believe That Guidance Is Better Than Answers This is one I have no problem with. I prefer guidance any time I am learning from another developer. Answers may get you going, but guidance will leave you stranded. At some point, you will come across a problem that can only be solved by thinking for yourself and this is where that guidance will really come in handy. You can use that guidance and extrapolate whatever technology to salve that problem (if it’s the right tool for solving that problem). The problem is, lots of developers simply want someone to tell them, “Do this, then this, then set that, and write this.” Favor thinking and learn the guidance of doing X and don’t ask someone to show you how to do X, if that makes sense. Read, Read and Read If you don’t like reading, you’re probably NOT going to make it into the Great Developer group. Great developers read books, they read magazines and they read code. Open source playgrounds like SourceForge, CodePlex and GitHub, have made it extremely easy to download code from developers you admire and see how they do stuff. Chances are, if you read their blog too, they’ll even explain WHY they did what they did (see “Guidance” above). MSDN and Code Magazine have not only code samples, but explanations of how to use certain technologies and sometimes even when NOT to use that same technology. Books are also out on just about every topic. I still favor the less technology centric books. For instance, I generally don’t buy books like, “Getting Started with Jiminy Jappets”. I look for titles like, “How To Write More Effective Code” (again, see guidance). The Addison-Wesley Signature Series is a great example of these types of books. They teach technology-agnostic concepts. Head First Design Patterns is another great guidance book. It teaches the "Gang Of Four" Design Patterns in a very easy-to-understand, picture-heavy way (I LIKE pictures). Hang Your Balls Out There Even though the advice came from a 3rd-shift Kinko’s attendant, doesn’t mean it’s not sound advice. Write some code and put it out for others to read, criticize and castigate you for. Understand that there are some real jerks out there who are absolute geniuses. Don’t be afraid to get some great advice wrapped in some really nasty language. Try to take what’s good about it and leave what’s not. I have a tough time with this myself. I don’t really have any code out there that is available for review (other than my demo code). It takes some guts to do, but in the end, there is no substitute for getting a community of developers to critique your code and give you ways to improve. Get Involved Speaking of community, the local and online user groups and discussion forums are a great place to hear about technologies and techniques you might never come across otherwise. Mostly because you might not know to look. But, once you sit down with a bunch of other developers and start discussing what you’re interested in, you may open up a whole new perspective on it. Don’t just go to the UG meetings and watch the presentations either, get out there and talk, socialize. I realize geeks weren’t meant to necessarily be social creatures, but if you’re amongst other geeks, it’s much easier. I’ve learned more in the last 3-4 years that I have been involved in the community that I did in my previous 8 years of coding without it. Socializing works, even if socialism doesn’t. Continuous Improvement Lean proponents might call this “Kaizen”, but I call it progress. We all know, especially in the technology realm, if you’re not moving ahead, you’re falling behind. It may seem like drinking from a fire hose, but step back and pick out the technologies that speak to you. The ones that may you’re little heart go pitter-patter. Concentrate on those. If you’re still overloaded, pick the best of the best. Just know that if you’re not looking at the code you wrote last week or at least last year with some embarrassment, you’re probably stagnating. That’s about all I can say about that, cause I am all out of clichés to throw at it. :0) Write Code Great painters paint, great writers write, and great developers write code. The most sure-fire way to improve your coding ability is to continue writing code. Don’t just write code that your work throws on you, pick that technology you love or are curious to know more about and walk through some blog demo examples. Take the language you use everyday and try to get it to do something crazy. Who knows, you might create the next Google search algorithm! All in all, being a great developer is about finding yourself in all this code. If it is just a job to you, you will probably never be one of the “Great Developers”, but you’re probably okay with that. If, on the other hand, you do aspire to greatness, get out there and GET it. No one’s going hand it to you.

  Read the article

• C# Open Source software that is useful for learning Design Patterns

  - by Fathom Savvy

  In college I took a class in Expert Systems. The language the book taught (CLIPS) was esoteric - Expert Systems: Principles and Programming, Fourth Edition. I remember having a tough time with it. So, after almost failing the class, I needed to create the most awesome Expert System for my final presentation. I chose to create an expert system that would calculate risk analysis for a person's retirement portfolio. In short, the system would provide the services normally performed by one's financial adviser. In other words, based on personality, age, state of the macro economy, and other factors, should one's portfolio be conservative, moderate, or aggressive? In the appendix of the book (or on the CD-ROM), there was this in-depth example program for something unrelated to my presentation. Over my break, I read and re-read every line of that program until I understood it to the letter. Even though it was unrelated, I learned more than I ever could by reading all of the chapters. My presentation turned out to be pretty damn good and I received praises from my professor and classmates. So, the moral of the story is..., by understanding other people's code, you can gain greater insight into a language/paradigm than by reading canonical examples. Still, to this day, I am having trouble with everyday design patterns such as the Factory Pattern. I would like to know if anyone could recommend open source software that would help me understand the Gang of Four design patterns, at the very least. I have read the books, but I'm having trouble writing code for the concepts in the real world. Perhaps, by studying code used in today's real world applications, it might just "click". I realize a piece of software may only implement one kind of design pattern. But, if the pattern is an implementation you think is good for learning, and you know what pattern to look for within the source, I'm hoping you can tell me about it. For example, the System.Linq.Expressions namespace has a good example of the Visitor Pattern. The client calls Expression.Accept(new ExpressionVisitor()), which calls ExpressionVisitor (VisitExtension), which calls back to Expression (VisitChildren), which then calls Expression (Accept) again - wooah, kinda convoluted. The point to note here is that VisitChildren is a virtual method. Both Expression and those classes derived from Expression can implement the VisitChildren method any way they want. This means that one type of Expression can run code that is completely different from another type of derived Expression, even though the ExpressionVisitor class is the same in the Accept method. (As a side note Expression.Accept is also virtual). In the end, the code provides a real world example that you won't get in any book because it's kinda confusing. To summarize, If you know of any open source software that uses a design pattern implementation you were impressed by, please list it here. I'm sure it will help many others besides just me. public class VisitorPatternTest { public void Main() { Expression normalExpr = new Expression(); normalExpr.Accept(new ExpressionVisitor()); Expression binExpr = new BinaryExpression(); binExpr.Accept(new ExpressionVisitor()); } } public class Expression { protected internal virtual Expression Accept(ExpressionVisitor visitor) { return visitor.VisitExtension(this); } protected internal virtual Expression VisitChildren(ExpressionVisitor visitor) { if (!this.CanReduce) { throw Error.MustBeReducible(); } return visitor.Visit(this.ReduceAndCheck()); } public virtual Expression Visit(Expression node) { if (node != null) { return node.Accept(this); } return null; } public Expression ReduceAndCheck() { if (!this.CanReduce) { throw Error.MustBeReducible(); } Expression expression = this.Reduce(); if ((expression == null) || (expression == this)) { throw Error.MustReduceToDifferent(); } if (!TypeUtils.AreReferenceAssignable(this.Type, expression.Type)) { throw Error.ReducedNotCompatible(); } return expression; } } public class BinaryExpression : Expression { protected internal override Expression Accept(ExpressionVisitor visitor) { return visitor.VisitBinary(this); } protected internal override Expression VisitChildren(ExpressionVisitor visitor) { return CreateDummyExpression(); } protected internal Expression CreateDummyExpression() { Expression dummy = new Expression(); return dummy; } } public class ExpressionVisitor { public virtual Expression Visit(Expression node) { if (node != null) { return node.Accept(this); } return null; } protected internal virtual Expression VisitExtension(Expression node) { return node.VisitChildren(this); } protected internal virtual Expression VisitBinary(BinaryExpression node) { return ValidateBinary(node, node.Update(this.Visit(node.Left), this.VisitAndConvert<LambdaExpression>(node.Conversion, "VisitBinary"), this.Visit(node.Right))); } }

  Read the article

• Install server 2012 on HP ML110 G7 with B110i controller, no disks found

  - by Molotch

  I'm trying to install Server 2012 on a HP ML110 G7 with a B110i controller and four non hot-swappable SATA drives. I just can't get the Server 2012 boot disk PE environment to find any disks. I have downloaded the latest SPP (Service Pack for Proliant 2012.10) and flashed the BIOS. I have tried two different HP drivers for B110i and Windows X64, 6.18.0.64 and 6.18.2.64 to no avail. I have tried setting the controller to both AHCI and legacy mode in the BIOS, no difference. HP:s SmartStart disc for G7 servers only support installation of up to Windows Server 2008R2. HP:s installation instructions for Server 2012 Essentials says boot from the windows disk and use the storage drivers found on the SPP (I can't find any storage drivers on the SPP disk).

  Read the article

• Can the Dell PowerEdge T320 boot from a USB stick in hard drive emulation mode?

  - by Icydog

  I am trying to install FreeNAS on a Dell PowerEdge T320. I've followed the instructions here (http://doc.freenas.org/index.php/Burning_an_IMG_File) to write the .img file to four different USB sticks with dd if=FreeNAS-9.2.1.5-RELEASE-x64.img of=/dev/sde, but I can't boot off of any of them. When I try to boot, I get the following screen: F1 FreeBSD F2 FreeBSD F3 Drive 0 Then it auto-selects F1 and either prints # about once a second forever, or it sits with a blinking cursor doing nothing. Forum posts and the wiki page linked above say to set the USB boot emulation mode to hard disk (as opposed to auto or CD/DVD/floppy), and older Dell PowerEdge models including the T310 have this option in the BIOS settings, but I cannot find it on the T320 with the latest BIOS 2.1.2. I have even tried writing the USB image to a USB hard disk, but with the same lack of success. Has anyone been able to find this USB boot setting on a T320, or been able to boot FreeNAS/FreeBSD from USB in some other way on a T320?

  Read the article

< Previous Page | 37 38 39 40 41 42 43 44 45 46 47 48  | Next Page >