Search Results

Search found 8048 results on 322 pages for 'partial upgrade'.

Page 91/322 | < Previous Page | 87 88 89 90 91 92 93 94 95 96 97 98  | Next Page >

  • node.js on CentOS box is at v0.6.18, yum doesn't update or upgrade it. Why?

    - by ariestav
    I'm currently working with a CentOS box that has a version of node installed, when I do: nodejs -v I get v0.6.18 But I noticed on nodejs.org website, that the latest release is 0.8.12, so do: sudo yum update nodejs I get Loaded plugins: fastestmirror Loading mirror speeds from cached hostfile * base: centos-mirror.jchost.net * epel: fedora-epel.mirror.lstn.net * extras: centos.mirror.lstn.net * updates: centos.mirror.lstn.net Setting up Update Process No Packages marked for Update What's the deal? Why doesn't yum find the latest version of node? Do I have to download the .tar.gz from nodejs.org and install it that way?

    Read the article

  • Hiding the Flash Message After a Time Delay

    - by Madhan ayyasamy
    Hi Friends,The flash hash is a great way to provide feedback to your users.Here is a quick tip for hiding the flash message after a period of time if you don’t want to leave it lingering around.First, add this line to the head of your layout to ensure the prototype and script.aculo.us javascript libraries are loaded:Next, add the following to either your layout (recommended), your view templates or a partial depending on your needs. I usually add this to a partial and include the partial in my layouts. "flash", :id = flash_type % "text/javascript" do % setTimeout("new Effect.Fade('');", 10000); This will wrap the flash message in a div with class=‘flash’ and id=‘error’, ‘notice’ or ‘warn’ depending on the flash key specified.The value ‘10000’ is the time in milliseconds before the flash will disappear. In this case, 10 seconds.This function looks pretty good and little javascript stunts like this can help make your site feel more professional. It’s also worth bearing in mind though, not everybody can see well or read as quickly as others so this may not be suitable for every application.Update:As Mitchell has pointed out (see comments below), it may be better to set the flash_type as the div class rather than it’s id. If there is the possibility that you’ll be showing more than one flash message per page, setting the flash_type as the div id will result in your HTML/XHTML code becoming invalid because the unique intentifier will be used more than once per page.Here is a slightly more complex version of the method shown above that will hide all divs with class ‘flash’ after a time delay, achieving the same effect and also ensuring your code stays valid with more than one flash message! "flash #{flash_type}" % "text/javascript" do % setTimeout("$$('div.flash').each(function(flash){ flash.hide();})", 10000); In this example, the div id is not set at all. Instead, each flash div will have class “div” and also class of the type of flash message (“error”, “warning” etc.).Have a Great Day..:)

    Read the article

  • How to build MVC Views that work with polymorphic domain model design?

    - by Johann de Swardt
    This is more of a "how would you do it" type of question. The application I'm working on is an ASP.NET MVC4 app using Razor syntax. I've got a nice domain model which has a few polymorphic classes, awesome to work with in the code, but I have a few questions regarding the MVC front-end. Views are easy to build for normal classes, but when it comes to the polymorphic ones I'm stuck on deciding how to implement them. The one (ugly) option is to build a page which handles the base type (eg. IContract) and has a bunch of if statements to check if we passed in a IServiceContract or ISupplyContract instance. Not pretty and very nasty to maintain. The other option is to build a view for each of these IContract child classes, breaking DRY principles completely. Don't like doing this for obvious reasons. Another option (also not great) is to split the view into chunks with partials and build partial views for each of the child types that are loaded into the main view for the base type, then deciding to show or hide the partial in a single if statement in the partial. Also messy. I've also been thinking about building a master page with sections for the fields that only occur in subclasses and to build views for each subclass referencing the master page. This looks like the least problematic solution? It will allow for fairly simple maintenance and it doesn't involve code duplication. What are your thoughts? Am I missing something obvious that will make our lives easier? Suggestions?

    Read the article

  • Silverlight Cream Top Posted Authors July to December, 2010

    - by Dave Campbell
    It's past the first of January, and it's now time to recognize devs that have a large number of posts in Silverlight Cream. Ground Rules I pick what posts are on the blog Only posts that go in the database are included The author has to appear in SC at least 4 of the 6 months considered I averaged the monthly posts and am only showing Authors with an average greater than 1. Here are the Top Posted Authors at Silverlight Cream for July 1, 2010 through December 31, 2010: It is my intention to post a new list sometime shortly after the 1st of every month to recognize the top posted in the previous 6 months, so next up is January 1! Some other metrics for Silverlight Cream: At the time of this posting there are 7304 articles aggregated and searchable by partial Author, partial Title, keywords (in the synopsis), or partial URL. There are also 118 tags by which the articles can be searched. This is an increase of 317 posts over last month. At the time of this posting there are 783 articles tagged wp7dev. This is an increase of 119 posts over last month, or over a third of the posts added. Stay in the 'Light!

    Read the article

  • Silverlight Cream Top Posted Authors August, 2010 to January, 2011

    - by Dave Campbell
    It's *way* past the first of February, and it's now time to recognize devs that have a large number of posts in Silverlight Cream. Ground Rules I pick what posts are on the blog Only posts that go in the database are included The author has to appear in SC at least 4 of the 6 months considered I averaged the monthly posts and am only showing Authors with an average greater than 1. Here are the Top Posted Authors at Silverlight Cream for August 1, 2010 through January 31, 2011: It is my intention to post a new list sometime shortly after the 1st of every month to recognize the top posted in the previous 6 months, so next up is March 1! Some other metrics for Silverlight Cream: At the time of this posting there are 7304 articles aggregated and searchable by partial Author, partial Title, keywords (in the synopsis), or partial URL. There are also 118 tags by which the articles can be searched. This is an increase of 265 posts over last month. At the time of this posting there are 783 articles tagged wp7dev. This is an increase of 155 posts over last month, or over half of the posts added. Stay in the 'Light!

    Read the article

  • Silverlight Cream Top Posted Authors September, 2010 to February, 2011

    - by Dave Campbell
    It's a bit past the first of March, and it's now time to recognize devs that have a large number of posts in Silverlight Cream. Ground Rules I pick what posts are on the blog Only posts that go in the database are included The author has to appear in SC at least 4 of the 6 months considered I averaged the monthly posts and am only showing Authors with an average greater than 1. Here are the Top Posted Authors at Silverlight Cream for September 1, 2010 through February 28, 2011: It is my intention to post a new list sometime shortly after the 1st of every month to recognize the top posted in the previous 6 months, so next up is March 1! Some other metrics for Silverlight Cream: At the time of this posting there are 7672 articles aggregated and searchable by partial Author, partial Title, keywords (in the synopsis), or partial URL. There are also 118 tags by which the articles can be searched. This is an increase of 368 posts over last month. At the time of this posting there are 984 articles tagged wp7dev. This is an increase of 201 posts over last month, or 54% of the posts added. Stay in the 'Light!

    Read the article

  • Ubuntu unattended-upgrades stops apache

    - by Robbie
    This morning i was alerted to the fact that both apache instances serving my app were not responding to requests from my load balancer. I attempted apachectl restart and it said apache was not running. So, i started apache on both instances and got the service up again. I then followed the logs and worked out that both had performed upgrades via the unattended-upgrades package moments before they stopped responding. /var/log/unattended-upgrades/unattended-upgrades.log 2013-07-02 06:30:51,875 INFO Starting unattended upgrades script 2013-07-02 06:30:51,875 INFO Allowed origins are: ['o=Ubuntu,a=precise-security'] 2013-07-02 06:33:57,771 INFO Packages that are upgraded: accountsservice apache2 apache2-mpm-prefork apache2-utils apache2.2-bin apache2.2-common apparmor apport apt apt-transport-https apt-utils bind9-host binutils dbus dnsutils gnupg gpgv isc-dhcp-client isc-dhcp-common krb5-locales libaccountsservice0 libapt-inst1.4 libapt-pkg4.12 libbind9-80 libc-bin libc-dev-bin libc6 libc6-dev libcurl3-gnutls libdbus-1-3 libdbus-glib-1-2 libdns81 libdrm-intel1 libdrm-nouveau1a libdrm-radeon1 libdrm2 libexpat1 libfreetype6 libgc1c2 libgnutls-dev libgnutls-openssl27 libgnutls26 libgnutlsxx27 libisc83 libisccc80 libisccfg82 liblwres80 libruby1.8 libx11-6 libx11-data libxcb1 libxext6 libxml2 linux-firmware linux-image-virtual linux-libc-dev linux-virtual multiarch-support openssl perl perl-base perl-modules python-apport python-crypto python-keyring python-problem-report python-software-properties ri1.8 ruby1.8 ruby1.8-dev sudo tzdata update-manager-core 2013-07-02 06:33:57,772 INFO Writing dpkg log to '/var/log/unattended-upgrades/unattended-upgrades-dpkg_2013-07-02_06:33:57.772399.log' 2013-07-02 06:36:10,584 INFO All upgrades installed I'm running Ubuntu 12.04 on Amazon EC2 servers. I have unattended-upgrades installed and configured as follows: /etc/apt/apt.conf.d/50unattended-upgrades // Automatically upgrade packages from these (origin:archive) pairs Unattended-Upgrade::Allowed-Origins { "${distro_id}:${distro_codename}-security"; // "${distro_id}:${distro_codename}-updates"; // "${distro_id}:${distro_codename}-proposed"; // "${distro_id}:${distro_codename}-backports"; }; // List of packages to not update Unattended-Upgrade::Package-Blacklist { }; /etc/apt/apt.conf.d/20auto-upgrades APT::Periodic::Update-Package-Lists "1"; APT::Periodic::Unattended-Upgrade "1"; I've struggled to find documentation about what happens to running processes during an upgrade. - Is this expected behaviour? Or should unattended-upgrades restart apache after upgrading it? - What can I do to ensure apache is restarted correctly? Should I just blacklist the apache package?

    Read the article

  • after return PartialView() Url.Actionlink("Action", "Controller"), the Controller is lost

    - by Johannes
    Well the Question is related to a problem I posted before (http://stackoverflow.com/questions/2403899/asp-net-mvc-partial-view-does-not-call-my-action). In practice I've a partial view which contains a Form, after submitting the Form the Controller returns the Partial View. Well the Problem is if I reload the page which contains the partial view the function <%= Url.Action("ChangePassword", "Account") %> returns "Account/ChangePassword", if I submit the form and the partial is returned by the controller. Using return PartialView() the function <%= Url.Action("ChangePassword", "Account") %> returns only "ChangePassword". Any Idea because? The View looks like: <form action="<%= Url.Action("ChangePassword", "Account") %>" method="post" id="jform"> <div> <fieldset> <legend>Account Information</legend> <p> <label for="currentPassword">Current password:</label> <%= Html.Password("currentPassword") %> <%= Html.ValidationMessage("currentPassword") %> </p> <p> <label for="newPassword">New password:</label> <%= Html.Password("newPassword") %> <%= Html.ValidationMessage("newPassword") %> </p> <p> <label for="confirmPassword">Confirm new password:</label> <%= Html.Password("confirmPassword") %> <%= Html.ValidationMessage("confirmPassword") %> </p> <p> <input type="submit" value="Change Password" /> </p> </fieldset> </div> </form> </div> <script> $(function() { $('#jform').submit(function() { $('#jform').ajaxSubmit({ target: '#FmChangePassword' }); return false; }); }); </script> Part of the Controller: if (!ValidateChangePassword(currentPassword, newPassword, confirmPassword)) { return PartialView(ViewData); }

    Read the article

  • ASP.NET MVC, Webform hybrid

    - by Greg Ogle
    We (me and my team) have a ASP.NET MVC application and we are integrating a page or two that are Web Forms. We are trying to reuse the Master Page from our MVC part of the app in the WebForms part. We have found a way of rendering an MVC partial view in web forms, which works great, until we try and do a postback, which is the reason for using a WebForm. The Error: Validation of viewstate MAC failed. If this application is hosted by a Web Farm or cluster, ensure that configuration specifies the same validationKey and validation algorithm. AutoGenerate cannot be used in a cluster. The Code to render the partial view from a WebForm (credited to "How to include a partial view inside a webform"): public static class WebFormMVCUtil { public static void RenderPartial(string partialName, object model) { //get a wrapper for the legacy WebForm context var httpCtx = new HttpContextWrapper(System.Web.HttpContext.Current); //create a mock route that points to the empty controller var rt = new RouteData(); rt.Values.Add("controller", "WebFormController"); //create a controller context for the route and http context var ctx = new ControllerContext( new RequestContext(httpCtx, rt), new WebFormController()); //find the partial view using the viewengine var view = ViewEngines.Engines.FindPartialView(ctx, partialName).View; //create a view context and assign the model var vctx = new ViewContext(ctx, view, new ViewDataDictionary { Model = model }, new TempDataDictionary()); //ERROR OCCURS ON THIS LINE view.Render(vctx, System.Web.HttpContext.Current.Response.Output); } } My only experience with this error is in context of a web farm, which is not the case. Also, I understand that the machine key is used for decrypting the ViewState. Any information on how to diagnose this issue would be appreciated. A Work-around: So far the work-around is to move the header content to a PartialView, then use an AJAX call to call a page with just the Partial View from the WebForms, and then using the PartialView directly on the MVC Views. Also, we are still able to share non-tech-specific parts of the Master Page, i.e. anything that is not MVC specific. Still yet, this is not an ideal solution, a server-side solution is still desired. Also, this solutino has issues when working with controls that have more sophisticated controls, using JavaScript, particularly dynamically generated script as used by 3rd party controls.

    Read the article

  • Why always fires OnFailure when return View() to Ajax Form ?

    - by Wahid Bitar
    I'm trying to make a log-in log-off with Ajax supported. I made some logic in my controller to sign the user in and then return simple partial containing welcome message and log-Off ActionLink my Action method looks like this : public ActionResult LogOn(LogOnModel model, string returnUrl) { if (ModelState.IsValid) { if (MembershipService.ValidateUser(model.UserName, model.Password)) { FormsService.SignIn(model.UserName, model.RememberMe); if (Request.IsAjaxRequest()) { //HERE IS THE PROBLEM :( return View("LogedInForm"); } else { if (!String.IsNullOrEmpty(returnUrl)) return Redirect(returnUrl); else return RedirectToAction("Index", "Home"); } } else { ModelState.AddModelError("", "The user name or password provided is incorrect."); if (Request.IsAjaxRequest()) { return Content("There were an error !"); } } } return View(model); } and I'm trying to return this simple partial : Welcome <b><%= Html.Encode(Model.UserName)%></b>! <%= Html.ActionLink("Log Off", "LogOff", "Account") %> and of-course the two partial are strongly-typed to LogOnModel. But if i returned View("PartialName") i always get OnFailure with status code 500. While if i returned Content("My Message") everything is going right. so please tell me why i always get this "StatusCode = 500" ??. where is the big mistake ??. By the way in my Site MasterPage i rendered partial to show long-on simple form this partial looks like this : <script type="text/javascript"> function ShowErrorMessage(ajaxContext) { var response = ajaxContext.get_response(); var statusCode = response.get_statusCode(); alert("Sorry, the request failed with status code " + statusCode); } function ShowSuccessMessage() { alert("Hey everything is OK!"); } </script> <div id="logedInDiv"> </div> <% using (Ajax.BeginForm("LogOn", "Account", new AjaxOptions { UpdateTargetId = "logedInDiv", InsertionMode = InsertionMode.Replace, OnSuccess = "ShowSuccessMessage", OnFailure = "ShowErrorMessage" })) { %> <%= Html.TextBoxFor(m => m.UserName)%> <%= Html.PasswordFor(m => m.Password)%> <%= Html.CheckBoxFor(m => m.RememberMe)%> <input type="submit" value="Log On" /> < <% } %>

    Read the article

  • Cast IEnumerable<Inherited> To IEnumerable<Base>

    - by david2342
    I'm trying to cast an IEnumerable of an inherited type to IEnumerable of base class. Have tried following: var test = resultFromDb.Cast<BookedResource>(); return test.ToList(); But getting error: You cannot convert these types. Linq to Entities only supports conversion primitive EDM-types. The classes involved look like this: public partial class HistoryBookedResource : BookedResource { } public partial class HistoryBookedResource { public int ResourceId { get; set; } public string DateFrom { get; set; } public string TimeFrom { get; set; } public string TimeTo { get; set; } } public partial class BookedResource { public int ResourceId { get; set; } public string DateFrom { get; set; } public string TimeFrom { get; set; } public string TimeTo { get; set; } } [MetadataType(typeof(BookedResourceMetaData))] public partial class BookedResource { } public class BookedResourceMetaData { [Required(ErrorMessage = "Resource id is Required")] [Range(0, int.MaxValue, ErrorMessage = "Resource id is must be an number")] public object ResourceId { get; set; } [Required(ErrorMessage = "Date is Required")] public object DateFrom { get; set; } [Required(ErrorMessage = "Time From is Required")] public object TimeFrom { get; set; } [Required(ErrorMessage = "Time to is Required")] public object TimeTo { get; set; } } The problem I'm trying to solve is to get records from table HistoryBookedResource and have the result in an IEnumerable<BookedResource> using Entity Framework and LINQ. UPDATE: When using the following the cast seams to work but when trying to loop with a foreach the data is lost. resultFromDb.ToList() as IEnumerable<BookedResource>; UPDATE 2: Im using entity frameworks generated model, model (edmx) is created from database, edmx include classes that reprecent the database tables. In database i have a history table for old BookedResource and it can happen that the user want to look at these and to get the old data from the database entity framework uses classes with the same name as the tables to receive data from db. So i receive the data from table HistoryBookedResource in HistoryBookedResource class. Because entity framework generate the partial classes with the properties i dont know if i can make them virtual and override. Any suggestions will be greatly appreciated.

    Read the article

  • Unusual RJS error

    - by rrb
    Hi, I am getting the following error in my RoR application: RJS error: TypeError: element is null Element.update("notice", "Comment Posted"); Element.update("allcomments", "\n\n\n \n\n waht now?\n\n \n\n \n\n \n\n asdfasdfa\n \n\n \n\n asdfasdf\n \n\n\n\n\n"); But when I hit the refresh button, I can see my partial updated. Here's my code: show_comments View: <table> <% comments.each do |my_comment| %> <tr> <td><%=h my_comment.comment%></td> </tr> <% end %> </table> show View: <div class="wrapper"> <div class="rescale"> <div class="img-main"> <%= image_tag @deal.photo.url %> </div> </div> <div class="description"> <p class ="description_content"> <%=h @deal.description %> </p> </div> </div> <p> <b>Category:</b> <%=h @deal.category %> </p> <p> <b>Base price:</b> <%=h @deal.base_price %> </p> <%#*<p>%> <%#*<b>Discount:</b>%> <%#=h @deal.discount %> <%#*</p>%> <%= link_to 'Edit', edit_deal_path(@deal) %> | <%= link_to 'Back', deals_path %> <p> <%= render :partial=>'deal_comments', :locals=>{ :comments=>Comment.new(:deal_id=>@deal.id)} %> </p> <div id="allcomments"> <%= render :partial=>'show_comments', :locals=>{ :comments=>Comment.find(@deal.comments)} %> </div> Controller: def create @comment = Comment.new(params[:comment]) render :update do |page| if @comment.save page.replace_html 'notice', 'Comment Posted' else page.replace_html 'notice', 'Something went wrong' end page.replace_html 'allcomments', :partial=> 'deals/show_comments', :locals=>{:comments=> @comment.deal.comments} end end def show_comments @deal = Deal.find(params[:deal_id]) render :partial=> "deals/show_comments", :locals=>{:comments=>@deal.comments} end end

    Read the article

  • Ajax page.replace_html problems with partials in Rails

    - by Chris Power
    Hello, I am having a problem with a pretty simple AJAX call in rails. I have a blog-style application and each post has a "like" feature. I want to be able to increment the "like" on each post in the index using AJAX onclick. I got it to work; however, the DOM is a bit tricky here, because no matter what partial its looking at, it will only update the TOP partial. so if I click "like" on post #2, it will update and replace the "likes" on post #1 instead. Code for _post partial: <some code here...> <div id="postcontent"> Posted <%= post.created_at.strftime("%A, %b %d")%> <br /> </div> <div id="postlikes"> <%= link_to_remote 'Like', :url => {:controller => 'posts', :action => 'like_post', :id => post.id}%> <%= post.like %> </div> code for _postlikes partial: <div id="postlikes"> <%= link_to_remote 'Like', :url => {:controller => 'posts', :action => 'like_post', :id => @post.id}%> <%= @post.like %> </div> </div> like_post.rjs code: page.replace_html "postlikes", :partial => "postlikes", :object => @post page.visual_effect :highlight, "postlikes", :duration => 3 So this all works properly for the first "postcontent" div. But this is an index of posts, so if I wanted to updated the second "postcontent" div on the page, it will still replace the html of the first. I understand the problem, I just don't know how to fix it :) Thanks in advance!

    Read the article

  • Improving Partitioned Table Join Performance

    - by Paul White
    The query optimizer does not always choose an optimal strategy when joining partitioned tables. This post looks at an example, showing how a manual rewrite of the query can almost double performance, while reducing the memory grant to almost nothing. Test Data The two tables in this example use a common partitioning partition scheme. The partition function uses 41 equal-size partitions: CREATE PARTITION FUNCTION PFT (integer) AS RANGE RIGHT FOR VALUES ( 125000, 250000, 375000, 500000, 625000, 750000, 875000, 1000000, 1125000, 1250000, 1375000, 1500000, 1625000, 1750000, 1875000, 2000000, 2125000, 2250000, 2375000, 2500000, 2625000, 2750000, 2875000, 3000000, 3125000, 3250000, 3375000, 3500000, 3625000, 3750000, 3875000, 4000000, 4125000, 4250000, 4375000, 4500000, 4625000, 4750000, 4875000, 5000000 ); GO CREATE PARTITION SCHEME PST AS PARTITION PFT ALL TO ([PRIMARY]); There two tables are: CREATE TABLE dbo.T1 ( TID integer NOT NULL IDENTITY(0,1), Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T1 PRIMARY KEY CLUSTERED (TID) ON PST (TID) );   CREATE TABLE dbo.T2 ( TID integer NOT NULL, Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T2 PRIMARY KEY CLUSTERED (TID, Column1) ON PST (TID) ); The next script loads 5 million rows into T1 with a pseudo-random value between 1 and 5 for Column1. The table is partitioned on the IDENTITY column TID: INSERT dbo.T1 WITH (TABLOCKX) (Column1) SELECT (ABS(CHECKSUM(NEWID())) % 5) + 1 FROM dbo.Numbers AS N WHERE n BETWEEN 1 AND 5000000; In case you don’t already have an auxiliary table of numbers lying around, here’s a script to create one with 10 million rows: CREATE TABLE dbo.Numbers (n bigint PRIMARY KEY);   WITH L0 AS(SELECT 1 AS c UNION ALL SELECT 1), L1 AS(SELECT 1 AS c FROM L0 AS A CROSS JOIN L0 AS B), L2 AS(SELECT 1 AS c FROM L1 AS A CROSS JOIN L1 AS B), L3 AS(SELECT 1 AS c FROM L2 AS A CROSS JOIN L2 AS B), L4 AS(SELECT 1 AS c FROM L3 AS A CROSS JOIN L3 AS B), L5 AS(SELECT 1 AS c FROM L4 AS A CROSS JOIN L4 AS B), Nums AS(SELECT ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) AS n FROM L5) INSERT dbo.Numbers WITH (TABLOCKX) SELECT TOP (10000000) n FROM Nums ORDER BY n OPTION (MAXDOP 1); Table T1 contains data like this: Next we load data into table T2. The relationship between the two tables is that table 2 contains ‘n’ rows for each row in table 1, where ‘n’ is determined by the value in Column1 of table T1. There is nothing particularly special about the data or distribution, by the way. INSERT dbo.T2 WITH (TABLOCKX) (TID, Column1) SELECT T.TID, N.n FROM dbo.T1 AS T JOIN dbo.Numbers AS N ON N.n >= 1 AND N.n <= T.Column1; Table T2 ends up containing about 15 million rows: The primary key for table T2 is a combination of TID and Column1. The data is partitioned according to the value in column TID alone. Partition Distribution The following query shows the number of rows in each partition of table T1: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T1 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are 40 partitions containing 125,000 rows (40 * 125k = 5m rows). The rightmost partition remains empty. The next query shows the distribution for table 2: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T2 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are roughly 375,000 rows in each partition (the rightmost partition is also empty): Ok, that’s the test data done. Test Query and Execution Plan The task is to count the rows resulting from joining tables 1 and 2 on the TID column: SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; The optimizer chooses a plan using parallel hash join, and partial aggregation: The Plan Explorer plan tree view shows accurate cardinality estimates and an even distribution of rows across threads (click to enlarge the image): With a warm data cache, the STATISTICS IO output shows that no physical I/O was needed, and all 41 partitions were touched: Running the query without actual execution plan or STATISTICS IO information for maximum performance, the query returns in around 2600ms. Execution Plan Analysis The first step toward improving on the execution plan produced by the query optimizer is to understand how it works, at least in outline. The two parallel Clustered Index Scans use multiple threads to read rows from tables T1 and T2. Parallel scan uses a demand-based scheme where threads are given page(s) to scan from the table as needed. This arrangement has certain important advantages, but does result in an unpredictable distribution of rows amongst threads. The point is that multiple threads cooperate to scan the whole table, but it is impossible to predict which rows end up on which threads. For correct results from the parallel hash join, the execution plan has to ensure that rows from T1 and T2 that might join are processed on the same thread. For example, if a row from T1 with join key value ‘1234’ is placed in thread 5’s hash table, the execution plan must guarantee that any rows from T2 that also have join key value ‘1234’ probe thread 5’s hash table for matches. The way this guarantee is enforced in this parallel hash join plan is by repartitioning rows to threads after each parallel scan. The two repartitioning exchanges route rows to threads using a hash function over the hash join keys. The two repartitioning exchanges use the same hash function so rows from T1 and T2 with the same join key must end up on the same hash join thread. Expensive Exchanges This business of repartitioning rows between threads can be very expensive, especially if a large number of rows is involved. The execution plan selected by the optimizer moves 5 million rows through one repartitioning exchange and around 15 million across the other. As a first step toward removing these exchanges, consider the execution plan selected by the optimizer if we join just one partition from each table, disallowing parallelism: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = 1 AND $PARTITION.PFT(T2.TID) = 1 OPTION (MAXDOP 1); The optimizer has chosen a (one-to-many) merge join instead of a hash join. The single-partition query completes in around 100ms. If everything scaled linearly, we would expect that extending this strategy to all 40 populated partitions would result in an execution time around 4000ms. Using parallelism could reduce that further, perhaps to be competitive with the parallel hash join chosen by the optimizer. This raises a question. If the most efficient way to join one partition from each of the tables is to use a merge join, why does the optimizer not choose a merge join for the full query? Forcing a Merge Join Let’s force the optimizer to use a merge join on the test query using a hint: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN); This is the execution plan selected by the optimizer: This plan results in the same number of logical reads reported previously, but instead of 2600ms the query takes 5000ms. The natural explanation for this drop in performance is that the merge join plan is only using a single thread, whereas the parallel hash join plan could use multiple threads. Parallel Merge Join We can get a parallel merge join plan using the same query hint as before, and adding trace flag 8649: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN, QUERYTRACEON 8649); The execution plan is: This looks promising. It uses a similar strategy to distribute work across threads as seen for the parallel hash join. In practice though, performance is disappointing. On a typical run, the parallel merge plan runs for around 8400ms; slower than the single-threaded merge join plan (5000ms) and much worse than the 2600ms for the parallel hash join. We seem to be going backwards! The logical reads for the parallel merge are still exactly the same as before, with no physical IOs. The cardinality estimates and thread distribution are also still very good (click to enlarge): A big clue to the reason for the poor performance is shown in the wait statistics (captured by Plan Explorer Pro): CXPACKET waits require careful interpretation, and are most often benign, but in this case excessive waiting occurs at the repartitioning exchanges. Unlike the parallel hash join, the repartitioning exchanges in this plan are order-preserving ‘merging’ exchanges (because merge join requires ordered inputs): Parallelism works best when threads can just grab any available unit of work and get on with processing it. Preserving order introduces inter-thread dependencies that can easily lead to significant waits occurring. In extreme cases, these dependencies can result in an intra-query deadlock, though the details of that will have to wait for another time to explore in detail. The potential for waits and deadlocks leads the query optimizer to cost parallel merge join relatively highly, especially as the degree of parallelism (DOP) increases. This high costing resulted in the optimizer choosing a serial merge join rather than parallel in this case. The test results certainly confirm its reasoning. Collocated Joins In SQL Server 2008 and later, the optimizer has another available strategy when joining tables that share a common partition scheme. This strategy is a collocated join, also known as as a per-partition join. It can be applied in both serial and parallel execution plans, though it is limited to 2-way joins in the current optimizer. Whether the optimizer chooses a collocated join or not depends on cost estimation. The primary benefits of a collocated join are that it eliminates an exchange and requires less memory, as we will see next. Costing and Plan Selection The query optimizer did consider a collocated join for our original query, but it was rejected on cost grounds. The parallel hash join with repartitioning exchanges appeared to be a cheaper option. There is no query hint to force a collocated join, so we have to mess with the costing framework to produce one for our test query. Pretending that IOs cost 50 times more than usual is enough to convince the optimizer to use collocated join with our test query: -- Pretend IOs are 50x cost temporarily DBCC SETIOWEIGHT(50);   -- Co-located hash join SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (RECOMPILE);   -- Reset IO costing DBCC SETIOWEIGHT(1); Collocated Join Plan The estimated execution plan for the collocated join is: The Constant Scan contains one row for each partition of the shared partitioning scheme, from 1 to 41. The hash repartitioning exchanges seen previously are replaced by a single Distribute Streams exchange using Demand partitioning. Demand partitioning means that the next partition id is given to the next parallel thread that asks for one. My test machine has eight logical processors, and all are available for SQL Server to use. As a result, there are eight threads in the single parallel branch in this plan, each processing one partition from each table at a time. Once a thread finishes processing a partition, it grabs a new partition number from the Distribute Streams exchange…and so on until all partitions have been processed. It is important to understand that the parallel scans in this plan are different from the parallel hash join plan. Although the scans have the same parallelism icon, tables T1 and T2 are not being co-operatively scanned by multiple threads in the same way. Each thread reads a single partition of T1 and performs a hash match join with the same partition from table T2. The properties of the two Clustered Index Scans show a Seek Predicate (unusual for a scan!) limiting the rows to a single partition: The crucial point is that the join between T1 and T2 is on TID, and TID is the partitioning column for both tables. A thread that processes partition ‘n’ is guaranteed to see all rows that can possibly join on TID for that partition. In addition, no other thread will see rows from that partition, so this removes the need for repartitioning exchanges. CPU and Memory Efficiency Improvements The collocated join has removed two expensive repartitioning exchanges and added a single exchange processing 41 rows (one for each partition id). Remember, the parallel hash join plan exchanges had to process 5 million and 15 million rows. The amount of processor time spent on exchanges will be much lower in the collocated join plan. In addition, the collocated join plan has a maximum of 8 threads processing single partitions at any one time. The 41 partitions will all be processed eventually, but a new partition is not started until a thread asks for it. Threads can reuse hash table memory for the new partition. The parallel hash join plan also had 8 hash tables, but with all 5,000,000 build rows loaded at the same time. The collocated plan needs memory for only 8 * 125,000 = 1,000,000 rows at any one time. Collocated Hash Join Performance The collated join plan has disappointing performance in this case. The query runs for around 25,300ms despite the same IO statistics as usual. This is much the worst result so far, so what went wrong? It turns out that cardinality estimation for the single partition scans of table T1 is slightly low. The properties of the Clustered Index Scan of T1 (graphic immediately above) show the estimation was for 121,951 rows. This is a small shortfall compared with the 125,000 rows actually encountered, but it was enough to cause the hash join to spill to physical tempdb: A level 1 spill doesn’t sound too bad, until you realize that the spill to tempdb probably occurs for each of the 41 partitions. As a side note, the cardinality estimation error is a little surprising because the system tables accurately show there are 125,000 rows in every partition of T1. Unfortunately, the optimizer uses regular column and index statistics to derive cardinality estimates here rather than system table information (e.g. sys.partitions). Collocated Merge Join We will never know how well the collocated parallel hash join plan might have worked without the cardinality estimation error (and the resulting 41 spills to tempdb) but we do know: Merge join does not require a memory grant; and Merge join was the optimizer’s preferred join option for a single partition join Putting this all together, what we would really like to see is the same collocated join strategy, but using merge join instead of hash join. Unfortunately, the current query optimizer cannot produce a collocated merge join; it only knows how to do collocated hash join. So where does this leave us? CROSS APPLY sys.partitions We can try to write our own collocated join query. We can use sys.partitions to find the partition numbers, and CROSS APPLY to get a count per partition, with a final step to sum the partial counts. The following query implements this idea: SELECT row_count = SUM(Subtotals.cnt) FROM ( -- Partition numbers SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1 ) AS P CROSS APPLY ( -- Count per collocated join SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals; The estimated plan is: The cardinality estimates aren’t all that good here, especially the estimate for the scan of the system table underlying the sys.partitions view. Nevertheless, the plan shape is heading toward where we would like to be. Each partition number from the system table results in a per-partition scan of T1 and T2, a one-to-many Merge Join, and a Stream Aggregate to compute the partial counts. The final Stream Aggregate just sums the partial counts. Execution time for this query is around 3,500ms, with the same IO statistics as always. This compares favourably with 5,000ms for the serial plan produced by the optimizer with the OPTION (MERGE JOIN) hint. This is another case of the sum of the parts being less than the whole – summing 41 partial counts from 41 single-partition merge joins is faster than a single merge join and count over all partitions. Even so, this single-threaded collocated merge join is not as quick as the original parallel hash join plan, which executed in 2,600ms. On the positive side, our collocated merge join uses only one logical processor and requires no memory grant. The parallel hash join plan used 16 threads and reserved 569 MB of memory:   Using a Temporary Table Our collocated merge join plan should benefit from parallelism. The reason parallelism is not being used is that the query references a system table. We can work around that by writing the partition numbers to a temporary table (or table variable): SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   CREATE TABLE #P ( partition_number integer PRIMARY KEY);   INSERT #P (partition_number) SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1;   SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals;   DROP TABLE #P;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; Using the temporary table adds a few logical reads, but the overall execution time is still around 3500ms, indistinguishable from the same query without the temporary table. The problem is that the query optimizer still doesn’t choose a parallel plan for this query, though the removal of the system table reference means that it could if it chose to: In fact the optimizer did enter the parallel plan phase of query optimization (running search 1 for a second time): Unfortunately, the parallel plan found seemed to be more expensive than the serial plan. This is a crazy result, caused by the optimizer’s cost model not reducing operator CPU costs on the inner side of a nested loops join. Don’t get me started on that, we’ll be here all night. In this plan, everything expensive happens on the inner side of a nested loops join. Without a CPU cost reduction to compensate for the added cost of exchange operators, candidate parallel plans always look more expensive to the optimizer than the equivalent serial plan. Parallel Collocated Merge Join We can produce the desired parallel plan using trace flag 8649 again: SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: One difference between this plan and the collocated hash join plan is that a Repartition Streams exchange operator is used instead of Distribute Streams. The effect is similar, though not quite identical. The Repartition uses round-robin partitioning, meaning the next partition id is pushed to the next thread in sequence. The Distribute Streams exchange seen earlier used Demand partitioning, meaning the next partition id is pulled across the exchange by the next thread that is ready for more work. There are subtle performance implications for each partitioning option, but going into that would again take us too far off the main point of this post. Performance The important thing is the performance of this parallel collocated merge join – just 1350ms on a typical run. The list below shows all the alternatives from this post (all timings include creation, population, and deletion of the temporary table where appropriate) from quickest to slowest: Collocated parallel merge join: 1350ms Parallel hash join: 2600ms Collocated serial merge join: 3500ms Serial merge join: 5000ms Parallel merge join: 8400ms Collated parallel hash join: 25,300ms (hash spill per partition) The parallel collocated merge join requires no memory grant (aside from a paltry 1.2MB used for exchange buffers). This plan uses 16 threads at DOP 8; but 8 of those are (rather pointlessly) allocated to the parallel scan of the temporary table. These are minor concerns, but it turns out there is a way to address them if it bothers you. Parallel Collocated Merge Join with Demand Partitioning This final tweak replaces the temporary table with a hard-coded list of partition ids (dynamic SQL could be used to generate this query from sys.partitions): SELECT row_count = SUM(Subtotals.cnt) FROM ( VALUES (1),(2),(3),(4),(5),(6),(7),(8),(9),(10), (11),(12),(13),(14),(15),(16),(17),(18),(19),(20), (21),(22),(23),(24),(25),(26),(27),(28),(29),(30), (31),(32),(33),(34),(35),(36),(37),(38),(39),(40),(41) ) AS P (partition_number) CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: The parallel collocated hash join plan is reproduced below for comparison: The manual rewrite has another advantage that has not been mentioned so far: the partial counts (per partition) can be computed earlier than the partial counts (per thread) in the optimizer’s collocated join plan. The earlier aggregation is performed by the extra Stream Aggregate under the nested loops join. The performance of the parallel collocated merge join is unchanged at around 1350ms. Final Words It is a shame that the current query optimizer does not consider a collocated merge join (Connect item closed as Won’t Fix). The example used in this post showed an improvement in execution time from 2600ms to 1350ms using a modestly-sized data set and limited parallelism. In addition, the memory requirement for the query was almost completely eliminated  – down from 569MB to 1.2MB. The problem with the parallel hash join selected by the optimizer is that it attempts to process the full data set all at once (albeit using eight threads). It requires a large memory grant to hold all 5 million rows from table T1 across the eight hash tables, and does not take advantage of the divide-and-conquer opportunity offered by the common partitioning. The great thing about the collocated join strategies is that each parallel thread works on a single partition from both tables, reading rows, performing the join, and computing a per-partition subtotal, before moving on to a new partition. From a thread’s point of view… If you have trouble visualizing what is happening from just looking at the parallel collocated merge join execution plan, let’s look at it again, but from the point of view of just one thread operating between the two Parallelism (exchange) operators. Our thread picks up a single partition id from the Distribute Streams exchange, and starts a merge join using ordered rows from partition 1 of table T1 and partition 1 of table T2. By definition, this is all happening on a single thread. As rows join, they are added to a (per-partition) count in the Stream Aggregate immediately above the Merge Join. Eventually, either T1 (partition 1) or T2 (partition 1) runs out of rows and the merge join stops. The per-partition count from the aggregate passes on through the Nested Loops join to another Stream Aggregate, which is maintaining a per-thread subtotal. Our same thread now picks up a new partition id from the exchange (say it gets id 9 this time). The count in the per-partition aggregate is reset to zero, and the processing of partition 9 of both tables proceeds just as it did for partition 1, and on the same thread. Each thread picks up a single partition id and processes all the data for that partition, completely independently from other threads working on other partitions. One thread might eventually process partitions (1, 9, 17, 25, 33, 41) while another is concurrently processing partitions (2, 10, 18, 26, 34) and so on for the other six threads at DOP 8. The point is that all 8 threads can execute independently and concurrently, continuing to process new partitions until the wider job (of which the thread has no knowledge!) is done. This divide-and-conquer technique can be much more efficient than simply splitting the entire workload across eight threads all at once. Related Reading Understanding and Using Parallelism in SQL Server Parallel Execution Plans Suck © 2013 Paul White – All Rights Reserved Twitter: @SQL_Kiwi

    Read the article

  • Cursors 1 Sets 0

    - by GrumpyOldDBA
    I had an interesting experience with a database I essentially know nothing about. On the server is a database which stores session state, Microsoft provide the code/database with their dot net, so I'm told. Anyway this database has sat happily on the production server for the past 4 years I guess, we've finally made the upgrade to SQL 2008 and the ASPState database has also been upgraded. It seems most likely that the performance increase of our upgrade tipped the usage of this database into...(read more)

    Read the article

  • Upgrading Apache to 2.2.23 on ubuntu 12.04 LTS

    - by Salil P
    We had done a PCI scan on one of our servers running Ubuntu 12.04 LTS with apache 2.2.22. The scan reported a vulnerability in apache 2.2.22 (Apache HTTP Server Zero-Length Directory Name in LD_LIBRARY_PATH Vulnerability).The report states to upgrade the version to the latest stable release of either 2.2.23 or 2.2.24. How do I upgrade to the 2.2.23 to fix the vulnerability or is there a patch available that can fix this and if yes can you let me know how that can be patched.

    Read the article

  • Mount external hdd in fluxbox ubuntu -12.04 commandline install

    - by jeroen
    I did the following: Install command line interface with ubuntu alternate install 12.04 in vmwareplayer5(9.2.2) After the base system was installed: sudo apt-get update, upgrade and dist-upgrade, sudo apt-get install xinit xorg fluxbox build-essential lxterminal gksu leafpad pcmanfm mc chromium-browser, this works. I also installed vmwaretools. My problem is being unable to mount any usb hdd or thumb drives. I'm new at building fluxbox so any help would be much appreciated!

    Read the article

  • Oracle Launches New Oracle Database 12c Administrator Certifications

    - by Brandye Barrington
    Today Oracle University announces the release of new Oracle Database 12c Administrator certifications. The new Oracle Database 12c certifications emphasize the foundational and advanced skills needed by Database Administrators and will prepare DBAs to leverage powerful new management and consolidation capabilities, resulting in an even more valuable credential for customers and partners. ORACLE CERTIFIED ASSOCIATE (OCA)  The Oracle Certified Associate (OCA) for Oracle Database 12c objectives measure IT professionals' mastery of day-to-day administration skills and their ability to manage the challenges they're likely to encounter on the job. This credential focuses on SQL skills, operational administration of the Oracle Database including performance and space management, and installing, patching and upgrading the Oracle Database. Earning the OCA credential requires successful completion of two exams: 1Z0-061 - Oracle Database 12c: SQL Fundamentals and 1Z0-062 - Oracle Database 12c: Installation and Administration. The OCA certification track also allows for several alternate exams which can be substituted for 1Z0-061. ORACLE CERTIFIED PROFESSIONAL (OCP) Building on the competencies in the Oracle Database 12c OCA certification, the Oracle Certified Professional (OCP) for Oracle Database 12c certification includes advanced knowledge and skills required of top-performing database administrators. The OCP credential focuses on developing and implementing backup and recovery strategies, designing consolidation strategies to exploit multitenant container and pluggable databases, and thorough understanding how CDB/PDBs fit into the DBaaS cloud-computing model. Today, Oracle is releasing 1Z0-060 - Upgrade to Oracle Database 12c, which allows Oracle Certified Professionals with credentials in Oracle 9i, Oracle Database 10g or Oracle Database 11g to upgrade to Oracle Database 12c with a single exam. The upgrade exam focuses on designing consolidation strategies to exploit multitenant container and pluggable databases, implementing Oracle 12c feature-rich ILM support, optimizing SQL execution using dynamic swapping of sub plans, implementing real-time data redaction within databases, as well as exploiting many additional performance, backup and recovery, security and partitioning enhancements. The exam also includes a thorough review of core DBA skills. Visit the OCP certification track for more details on the new upgrade exam as well as alternate certification paths. ORACLE CERTIFIED MASTER (OCM) The Oracle Certified Master (OCM) for Oracle Database 12c - a very challenging and elite top-level certification - certifies the most highly skilled and experienced database experts. Further information on the 12c OCM level will be announced as exam development concludes. To date, there have been more than 1.6 million Oracle certifications granted worldwide. Explore these certification tracks, exam requirements and objectives, and start toward earning your exciting new Oracle Database 12c certification credentials from Oracle.

    Read the article

  • No updates in my Raring

    - by zatloukal-frantisek
    Since upgrade from Quantal to raring i am not recieving any updates. For example firefox package - I have version 17 installed and apt-get update && apt-get upgrade does not find updates. And output from show-versions: fanys@fanys-netbook:~$ apt-show-versions firefox firefox 17.0+build2-0ubuntu0.12.10.1 newer than version in archive fanys@fanys-netbook:~$ apt-show-versions unity unity/raring uptodate 6.12.0-0ubuntu1 I tried to remove contents of /var/lib/apt/lists/ and redo package refresh(apt-get update). But still same issue. /etc/apt/sources.list contents: # See http://help.ubuntu.com/community/UpgradeNotes for how to upgrade to # newer versions of the distribution. deb http://cz.archive.ubuntu.com/ubuntu/ raring main restricted deb-src http://cz.archive.ubuntu.com/ubuntu/ raring main restricted ## Major bug fix updates produced after the final release of the ## distribution. deb http://cz.archive.ubuntu.com/ubuntu/ raring-updates main restricted deb-src http://cz.archive.ubuntu.com/ubuntu/ raring-updates main restricted ## N.B. software from this repository is ENTIRELY UNSUPPORTED by the Ubuntu ## team. Also, please note that software in universe WILL NOT receive any ## review or updates from the Ubuntu security team. deb http://cz.archive.ubuntu.com/ubuntu/ raring universe deb-src http://cz.archive.ubuntu.com/ubuntu/ raring universe deb http://cz.archive.ubuntu.com/ubuntu/ raring-updates universe deb-src http://cz.archive.ubuntu.com/ubuntu/ raring-updates universe ## N.B. software from this repository is ENTIRELY UNSUPPORTED by the Ubuntu ## team, and may not be under a free licence. Please satisfy yourself as to ## your rights to use the software. Also, please note that software in ## multiverse WILL NOT receive any review or updates from the Ubuntu ## security team. deb http://cz.archive.ubuntu.com/ubuntu/ raring multiverse deb-src http://cz.archive.ubuntu.com/ubuntu/ raring multiverse deb http://cz.archive.ubuntu.com/ubuntu/ raring-updates multiverse deb-src http://cz.archive.ubuntu.com/ubuntu/ raring-updates multiverse ## N.B. software from this repository may not have been tested as ## extensively as that contained in the main release, although it includes ## newer versions of some applications which may provide useful features. ## Also, please note that software in backports WILL NOT receive any review ## or updates from the Ubuntu security team. deb http://security.ubuntu.com/ubuntu raring-security main restricted deb-src http://security.ubuntu.com/ubuntu raring-security main restricted deb http://security.ubuntu.com/ubuntu raring-security universe deb-src http://security.ubuntu.com/ubuntu raring-security universe deb http://security.ubuntu.com/ubuntu raring-security multiverse deb-src http://security.ubuntu.com/ubuntu raring-security multiverse ## Uncomment the following two lines to add software from Canonical's ## 'partner' repository. ## This software is not part of Ubuntu, but is offered by Canonical and the ## respective vendors as a service to Ubuntu users. deb http://archive.canonical.com/ubuntu raring partner deb-src http://archive.canonical.com/ubuntu raring partner ## This software is not part of Ubuntu, but is offered by third-party ## developers who want to ship their latest software. deb http://extras.ubuntu.com/ubuntu raring main deb-src http://extras.ubuntu.com/ubuntu raring main deb http://cz.archive.ubuntu.com/ubuntu/ raring-proposed main universe restricted multiverse deb http://cz.archive.ubuntu.com/ubuntu/ raring-backports main universe restricted multiverse I have no updates for 4 days of dist-upgrade. There is one package kept in actual version: libexttextcat-data Thanks in advance

    Read the article

  • OBIEE 11.1.1 - BI Design Best Practices Whitepaper V1.2

    - by Nicolas Barasz
    Oracle BI Principles. Repository design best practices. Dashboards and reports design best practices. 10g Upgrade considerations. This new version includes 40 more slides than the previous one. Multiple new best practices specific to 11g and a lot of new information about upgrade from 10g. Click here to download (Right click or option-click the link and choose "Save As..." to download this pdf file)

    Read the article

  • Achieving Zero Downtime Deployment

    - by MattW
    I am trying to achieve zero downtime deployments so I can deploy less during off hours and more during "slower" hours - or anytime, in theory. My current setup, somewhat simplified: Web Server A (.NET App) Web Server B (.NET App) Database Server (SQL Server) My current deployment process: "Stop" the sites on both Web Server A and B Upgrade the database schema for the version of the app being deployed Update Web Server A Update Web Server B Bring everything back online Current Problem This leads to a small amount of downtime each month - about 30 mins. I do this during off hours, so it isn't a huge problem - but it is something I'd like to get away from. Also - there is no way to really go 'back'. I don't generally make rollback DB scripts - only upgrade scripts. Leveraging The Load Balancer I'd love to be able to upgrade one Web Server at a time. Take Web Server A out of the load balancer, upgrade it, put it back online, then repeat for Web Server B. The problem is the database. Each version of my software will need to execute against a different version of the database - so I am sort of "stuck". Possible Solution A current solution I am considering is adopting the following rules: Never delete a database table. Never delete a database column. Never rename a database column. Never reorder a column. Every stored procedure must be versioned. Meaning - 'spFindAllThings' will become 'spFindAllThings_2' when it is edited. Then it becomes 'spFindAllThings_3' when edited again. Same rule applies to views. While, this seems a bit extreme - I think it solves the problem. Each version of the application will be hitting the DB in a non breaking way. The code expects certain results from the views/stored procedures - and this keeps that 'contract' valid. The problem is - it just seeps sloppy. I know I can clean up old stored procedures after the app is deployed for awhile, but it just feels dirty. Also - it depends on all of the developers following these rule, which will mostly happen, but I imagine someone will make a mistake. Finally - My Question Is this sloppy or hacky? Is anybody else doing it this way? How are other people solving this problem?

    Read the article

  • Godaddy VPS or swith to another provider?

    - by Charlie
    Long story short, I want to be able to have close to full control over my server, mainly being able to install things on my sever that Godaddy shared hosting does not allow (gzip, etc..). Should I switch providers to something else (all my domains are hosted there) or upgrade my hosting to VPS. If I upgrade, how hard will it be to set it up without their "assisted service" or something setup (where they do virus scanning, etc.)?

    Read the article

  • Oracle Support Customers take note My Oracle Support Flash is set to Retire

    - by user12244613
    Take Action – My Oracle Support Flash User Interface Set to Retire On July 13, 2012, Oracle plans to upgrade the HTML interface with additional functionality that will allow those users still remaining on the Flash-based interface to switch over to the HTML version. Although the Flash-based user interface will remain available for a brief period following the upgrade, we encourage you to begin using the new HTML version sooner. Find out when you should make the switch! Read complete communication to Flash users

    Read the article

  • LTS vs normal release software versions synced from Debian

    - by Jasper Loy
    I read that LTS releases are based on Debian testing while normal releases are based on Debian unstable. Given the long release cycle of Debian, is it possible for some software to be of a more recent version in a normal release X than in LTS release X+1? If the answer is yes, would there be a difference between an upgrade and a fresh install (perhaps upgrade holds back more recent version automatically)?

    Read the article

  • Don't Miss Oracle UPK at the Oracle Applications Virtual Tradeshow

    - by di.seghposs(at)oracle.com
    Be sure to visit the Oracle Applications Virtual Tradeshow - Spotlight on Customer Success - February 3, 2011. If you are considering using Oracle UPK for a project or an upgrade, this is an event you don't want to miss. Hear how the City and County of San Francisco used Oracle UPK for their successful PeopleSoft upgrade. Get a chance to meet the experts and listen to 20+ customers share their success with Oracle Applications. Register Now!

    Read the article

< Previous Page | 87 88 89 90 91 92 93 94 95 96 97 98  | Next Page >