Chris May

As far as I could find, there is no easy way to disable (not delete) an SQL Server Reporting Subscription.

We thought we had found a way to do this by changing the schedule to "run once" and set an "End Date" so that it would never run after that date.

This seemed to work, but after a restart of the server this weekend, the job executed on monday morning.

I'm not sure why it happened Monday morning ( as opposed to Sunday morning, as the machine was restarted on Saturday night) or if has anything to do with today being the first of the month, but either way, this didn't work and people were emailed some useless notice that they didn't need.

So, if you are using this method to disable your SSRS reports, watch out.

I had recently tried to use my Windows Home Server to restore a laptop image to a VM running inside HyperV. 

In the event that my machine totally died (not just a hard drive issue) I could restore the image to a VM and work from there while I got back on my feet with a new laptop.

Unfortunately I didn't find any way to make this work.  I ended up doing all this work and in the end, I couldn't boot.

A disk read error occurred
Press ctrl+alt+delete to restart.

I tried everything I could think of, repairing the MBR, running fixboot, doing a windows repair, restoring the hal dll from i386 etc.  Nothing worked.

So I gave up on using WHS to create an instance of my dev machine, and fell back on just trying to do a P2V of any type.

I came across Disk2VHD, a free Systernals tool.  This tool includes the options to run on the source machine while online, and also provides an option to "Fix" the HAL (hardware abstraction layer).  I think this was part of my original problem (the HAL issue).  My computer was setup to use SATA drivers, and HyperV only supports virtual IDE drivers.  I think (but am not sure) that this is what was causing my problem w/ using the WHS restore.

So anyway, Disk2VHD runs super fast, and pretty soon I had a VHD to try out. 

Now, this VHD didn't work either, but it would get me further along.  I would see the windows splash screen before the black screen of death.

What followed was a weird set of steps trying to figure out why things wouldn't work, but in the end, I never ended up needing to disable anything.... so I'm just going to try to write down the steps I took as best I can.

1. After the first BSOD I booted to safe mode (using F8).
2. Safe mode worked, and it seemed to find some new hardware, as I would have expected, but I don't think it found any of the drivers it needed (so maybe I had ejected the virtual windows CD or something).
3. Booting normally gave the black screen of death again.
4. I then found this page with instructions on how to replace the HAL.  Really these instructions were how to do Disk2VHD to a working VM, but I had already done most of the steps except the hal replacement.  I'm going to replace those instructions below:
1. 1)Downloaded Disk2VHD (v 1.3) from sysinternals (free) and unzipped on the laptop
   2)Find a place to store the new vhd - I used a 320GB external USB, but any storage device that pc can see and is big enough to hold your hard disk should work
   3)Run Disk2VHD and input the name and location for the new file
   4)On your virtual host create a new VPC (I use Virtual PC 2007) and add that new vhd you just created as the primary hard disk.
   5)Make sure you have an ISO of WinXP or an actual disk. If using an iso you'll need to mount it so it can be booted from.
   6)Fire up the new vpc and have it boot from that XP disc.
   7)Let it run setup and then select option 2 (R) to get repair console
   Login in to windows (need admin pw)
   9)Type “cd system32” press enter
   10)Assuming your CD (or mounted iso) is drive D - Type “expand d:\i386\halacpi.dl_” press enter
   11)Type “copy halacpi.dll hal.dll” press enter and then press y
   12)Reboot
   13)Press F8 to get to safe mode
   14)After the message that new stuff was installed, click OK for reboot
   15)After system boots back up, log in and install VM Additions
5. IIRC I still got the black screen of death.  It's possible that at this point behind the scenes it was doing a check of my disck or something, but I can't tell in the VM if the HDD is going or not so I just assumed it was locked.
6. I booted to safe mode and used MSconfig to disable everything, and also used /SOS, and some other options.
7. What followed was just a series of successful boots with me enabling more and more until I had everything enabled and it all still worked, so I'm not sure why I was getting the BlackSOD before I started this set of SafeMode boots.
8. Another thing I did at some point (maybe more than once) was try to install the VM Integration Services, which says that it's updating the HAL, so maybe that had something to do with it too.

The next 2 things that I'm going to try are, running a WHS restore to a VM that is already working (I don't think this will matter, but we'll see), and 2 trying to use the Disk2VHD on a machine that has a VHD mounted as a secondary drive.  I don't konw if this is even possible, but I was thinking that I could use WHS to create the VHD, mount the VHD as X: on some machine, and then tell Disk2VHD to create a VHD from X: and instruct it to to the "Fix HAL" thing.  I'm guessing that will not work as well, but we'll see.

 

The amount of time that this cost me is astounding for such a stupid issue.  (And, I'm going to lay the blame fully on whoever wrote the dbRoot object in the MyGeneration project.  Total improper use of properties).

The situation was this: I was trying to dynamically wire up all the plumbing that the mygeneration windows app does for you manually from the ZeusCmd command line.

Doing this would allow me to switch from DB to DB within my code gen application.

Because ZeusCmd runs first (it references my app and passes in what is needed), if you had previously saved a connection in your MyGeneration app, then this connection is used when ZeusCmd runs.  I didn't want this to happen because as I move from location to location different DB servers are available or not available, so if the last DB server I connected to was at office A, I didn't want my code gen crashing onload when I am out of said office.

So I ran MyGeneration and changed my connection to be <None> and saved the settings.  Then in my code gen, I tried to manually create the dbRoot object (named meta here) that was previously created for me my ZeusCmd and my saved MyGeneration settings.

Code like this:

meta.Connect("SQL", "Provider=SQLNCLI;Server=cmayt61p;Database=DBName;Trusted_Connection=yes;")

Seemed to do the trick.

However, my codegen started blowing up on my later in the process.  After tracking down the error I find that the mygeneration result column object can't figure out the .net type that would be the same as a sql server "int".  Strange, never had these problems before.

A lot of debugging leads me to find that:

meta.DbTarget
meta.Language 

Are both returning "" when before they were returning "SqlClient" and "VB.NET" when I was not dynamically connecting.

So here is where things get bad... anytime I try setting these properties to their values:

meta.DbTarget = "SqlClient"
meta.Language = "VB.NET"

Nothing happens.  The values don't appear to stick.

Eventually I even start trying to set the LanguageMappingFileName and DbTargetMappingFileName properties as well:

meta.DbTarget = "SqlClient"
meta.Language = "VB.NET"
meta.LanguageMappingFileName = "C:\Program Files\MyGeneration13\Settings\Languages.xml"
meta.DbTargetMappingFileName = "C:\Program Files\MyGeneration13\Settings\DbTargets.xml"

I tried doing these property sets before connecting and after connecting.

I tried returning to having MyGeneration setup the meta object form me (and all these properties have their values) before dynamically connecting (and losing the DbTarget and Language property values).

I even tried creating a brand new dbRoot object, as well as trying to read some hidden values in the IZeusInput object.

But, in the end, it all came down to the this 1 little thing:

You have to set the mapping file properties before you can set the DbTarget and Language properties.

So this will work just fine:

meta.Connect("SQL", "Provider=SQLNCLI;Server=cmayt61p;Database=MSiUpgraded;Trusted_Connection=yes;")
meta.LanguageMappingFileName = "C:\Program Files\MyGeneration13\Settings\Languages.xml"
meta.DbTargetMappingFileName = "C:\Program Files\MyGeneration13\Settings\DbTargets.xml"
meta.DbTarget = "SqlClient"
meta.Language = "VB.NET"

But this won't:

meta.Connect("SQL", "Provider=SQLNCLI;Server=cmayt61p;Database=MSiUpgraded;Trusted_Connection=yes;")
meta.DbTarget = "SqlClient"
meta.Language = "VB.NET"
meta.LanguageMappingFileName = "C:\Program Files\MyGeneration13\Settings\Languages.xml"
meta.DbTargetMappingFileName = "C:\Program Files\MyGeneration13\Settings\DbTargets.xml"

It just so happened that everywhere that I was testing this stuff in my code, I always set the DbTarget and Language properties first, and they would never stick.

This is not a proper use of properties.  If a property set has conditional logic that will prevent it from setting, there is something wrong here.  At minimum something should happen, an exception being thrown maybe?  The proper thing to do in my mind would be to make DbTarget and Language readonly properties and have functions to set their value, which could return something indicating failure (or throw an exception).

Property gets and sets should do just that, get and set properties.  There shouldn't be logic in there deciding if and when to actually perform the set that you have coded.  That is the job of a function.

UPDATE: And I discovered as well that you have to connect first before you set any of these values, or none of them will stick.  So you have to 1) connect 2) set the file paths 3) set the language and dbtargets, in that order or nothing works.

This page has a quick demo of all the easing methods in the jQuery Easing plugin:  http://www.commadot.com/jquery/easing.php

Pretty cool.  Just click on any of the boxes to see the easing action.

I was trying to use some jQuery effects on this page:

http://docs.jquery.com/UI/Effects

And none of them were working.  After a while, I realized it was because I didn't have the "jQuery Effects Core" or as they list it on that page: effects.core.js

Turns out you need to download it as part of the UI package here: http://jqueryui.com/download

About a year ago, we tried using a tool to get our SQL schema into a souce control system.  It was supposed to work like this: every night the process would run and checkin each object as a file into TFS. 

Well, aside from not working, it also added Extended Properties to all our objects in SQL Server. 

Not really a big deal, except when we would run our database comparison tools between production and development these would cause it so that almost everything would appear out of sync.

I ended up writing my own service to check in all our SQL objects into TFS, but we were left with all these Extended Properties all over the place. 

So I finally got around to cleaning them up using the script below:

select 'EXEC sp_dropextendedproperty 
@name = ''' + sys.extended_properties.name + ''' 
,@level0type = ''schema'' 
,@level0name = ' + object_schema_name(extended_properties.major_id) + '
,@level1type = ''' + 
    CASE WHEN  
        xtype = 'U' THEN 'table'
        WHEN 
        xtype = 'FN' THEN 'function'
        WHEN 
        xtype = 'V' THEN 'view'
        WHEN 
        xtype = 'P' THEN 'procedure' end
    + '''
,@level1name = ''' + object_name(extended_properties.major_id) + ''''
from sys.extended_properties
INNER JOIN sysobjects ON sys.extended_properties.major_id  = sysobjects.id
where extended_properties.class_desc = 'OBJECT_OR_COLUMN'
and extended_properties.minor_id = 0
and extended_properties.name like 'VSS%'

This script will produce a record set that you can copy and paste into another query window to run to clean out all the extended properties. 

Note: all the generated extended properties that this tool created on our DB started with "VSS" so that is why I have that extra item in my where clause.  You'll probably want to remove that.

Here is the article on how to install the update on the EX47X machines.

When HP released the HP MediaSmart Server 2.5 Update for the EX48x machines, they also ^[1] announced the full software would not be available to first generation users due "to hardware differences" anhd "the underlying software architecture". Based on that statement, Nigel Wilks (Cougar) and Alex Kuretz (Yakuza) started work on a way^[2] to make the update available to EX47x owners, dubbed "SanEncore" after the code names for the EX47x (San Juan) and the EX48x (Encore).

Cougar and Yakuza have also developed a Windows Home Server Add-In to fake the Update package into thinking it is running on a 2.1 machine, and handles some of the missing configuration items so that the 2.5 update is sucesful.

From this post:
http://forums.lhotka.net/forums/thread/25658.aspx


Child pattern:

[Serializable]
public class Child : BusinessBase<Child>
{
  internal static Child NewChild()
  {
    return DataPortal.CreateChild<Child>();
  }

  internal static Child GetChild()
  {
    return DataPortal.FetchChild<Child>();
  }

  private Child()
  { 
    MarkAsChild();
  }

  private void Child_Create()
  {
    // initialize new child here
  }

  private void Child_Fetch()
  {
    // load child data here
  }

  private void Child_Insert()
  {
    // insert child data here
  }

  private void Child_Update()
  {
    // update child data here
  }

  private void Child_DeleteSelf()
  {
    // delete child data here
  }
}

Editable root pattern:

[Serializable]
public class RootParent : BusinessBase<RootParent>
{
  // other class code here ...

  protected override void DataPortal_Insert()
  {
    using (SqlConnection cn = new SqlConnection(...))
    {
      // insert parent data here

      FieldManager.UpdateChildren();
    }
  }

  protected override void DataPortal_Update()
  {
    using (SqlConnection cn = new SqlConnection(...))
    {
      // update parent data here

      FieldManager.UpdateChildren();
    }
  }
}

Editable Root List:

[Serializable]
public class ChildList : BusinessListBase<ChildList, Child>
{
  // other class code here ...

  protected override void DataPortal_Update()
  {
    using (SqlConnection cn = new SqlConnection(...))
    {
      Child_Update();
    }
  }
}

Editable Child List:

[Serializable]
public class ChildList : BusinessListBase<ChildList, Child>
{
  internal ChildList()
  {
    MarkAsChild();
  }

  // other class code here ...
}

If you use the DataPortal methods for creating the child objects (i.e. FetchChild) then you don't have to bother with the MarkAsChild() stuff.

From time to time you might have an enum that you want to databind to something as if it were a list, so I wrote some classes that turn an Enum into a list that you can databind. 

I created ways to do this in CSLA and also w/o CSLA.  I made the non-CSLA versions 1 work so that you can either use a sub class, or just use generics when defining the type of your list.  You can do the same with the CSLA versions if you want by changing some of the generics, but I didn't bother.

'*** csla
Dim cslaList As MyTestTypeList = MyTestTypeList.GetList

'*** non-csla with inherited subclass and without
Dim list As New MyTestTypeList2
Dim list2 As New EnumList2(Of MyTestType)

'*** enum we are using
Public Enum MyTestType
    something = 1
    here = 2
    ok = 3
End Enum

'*** csla
<Serializable()> _
Public Class MyTestTypeList
    Inherits EnumList(Of mytesttype, MyTestTypeList)

End Class
<Serializable()> _
Public Class EnumList(Of T As Structure, R As EnumList(Of T, R))
    Inherits Csla.NameValueListBase(Of String, Integer)

    Public Shared Function GetList() As R
        Return Csla.DataPortal.Fetch(Of R)()
    End Function

    Protected Sub New()
        ' require use of factory method 
    End Sub

    <Csla.RunLocal()> _
    Public Overloads Sub DataPortal_Fetch() 'ByVal criteria As Object)
        Me.IsReadOnly = False
        For Each item As T In [Enum].GetValues(GetType(T))
            Dim name As String = [Enum].GetName(GetType(T), item)
            Add(New Csla.NameValueListBase(Of String, Integer).NameValuePair(name, [Enum].Parse(GetType(T), name)))
        Next
        Me.IsReadOnly = True
    End Sub 
End Class

'*** non-csla
Public Class MyTestTypeList2
    Inherits EnumList2(Of MyTestType)

End Class

Public Class EnumList2(Of T As Structure)
    Inherits List(Of EnumListItem)

    Public Structure EnumListItem
        Dim Name As String
        Dim Value As Integer
        Public Sub New(ByVal name As String, ByVal value As Integer)
            Me.Name = name
            Me.Value = value
        End Sub
    End Structure 

    Public Sub New()
        For Each item As T In [Enum].GetValues(GetType(T))
            Dim name As String = [Enum].GetName(GetType(T), item)
            Add(New EnumListItem(name, [Enum].Parse(GetType(T), name)))
        Next
    End Sub

End Class

This chat takes place after I have already chatted with a different support group, called the IBM parts divions, and the IBM tech support division:

Mohammed: Thank you for accepting our chat service, how may I assist you?
you: When I customize an x3500, at last screen it says: (5652) Planar Base [$699.00] is required for this configuration and has been added for you.
you: What is Planar Base and what about my configuration makes it a requirement?
Mohammed: Is this for yourself (your company) or a client of yours?
you: mysefl
Mohammed: I see. I'm not too familiar with the planar base. I can have a specialist contact you to further discuss this need of yours
you: ok. How long does that normally take?
Mohammed: our usual turnaround is 48 hours
Mohammed: however we can try to have somebody contact you by end of day tomorrow
Mohammed: but I can't make any promises
you: ok my number is (630) 708-0629
Mohammed: May I have your full name, telephone number, email address, job title, company name and physical address?
you: Chris May, (630) 708-0629, cmay@RealNameRemoved.com that should be good
Mohammed: unfortunately without all the information requested above, I will not be able to have a specialist contact you
you: ok forget it I'll just buy a Dell

UPDATE: I did buy a Dell.  Stupid IBM couldn't have someone take 5 minutes and answer a question for someone who wanted to give them $3000. 

When you login to your computer as Whatever\Username, you can have windows store usernames and passwords for Something\Othername and Hello\World using the Stored Usernames and Passwords.

Getting to this dialog can be a real pain.  Just run this command from the Run dialog.

rundll32.exe keymgr.dll,KRShowKeyMgr

 

Today I was trying to do a backup to my WHS and I was getting the message on my client:

Backup is waiting on (cleanup).

After quite a long time, it was still on this message.

I did some investigating on the server and saw that 45 minutes earlier my server had kicked off a backup cleanup process.  This was indicated in the event viewer under "Homeserver". 

Task Manager also showed whsbackup.exe doing a lot of work.

So, if you run into this, just wait a bit and when the server is done with the cleanup the client backup will start like normal.

So today I went to make some firewall updates for a client and the Cisco PDM wouldn't launch from the browser.

After some troubleshooting, I found that the Java VM was indicating an ArrayIndexOutofBoundsException had occurred.

After some checking around I confirmed my suspicion: Java sucks.  Just kidding, well not really, but what I really confirmed was that the PDM wouldn't work with any new version of Sun Java.  I guess I'm spoiled with .Net being backward compatible. 

Some suggested installing an old version of Java
http://java.sun.com/products/archive/j2se/1.4.2_03/index.html

But lucky for me I was able to just install Java 6 Update 15 and Java 6 Update 7 from Add/Remove programs and everything started working again.

It's totally true when people accuse MS of copying Java with the .NET framework, but they sure didn't make it suck like Java.

UPDATE: The version of Java that is working for me is Java 6 Update 6.  You can download it here:
http://java.sun.com/products/archive/j2se/6u6/index.html

I've been working on a project where I wanted to have a flyout window on the left just like how Visual Studio does their menus.

Maybe "slide out" is more accurate.

I used jQuery, which I am trying to use more in my projects, for the effects.

Anyway, I ended up making it a bit harder than it needed to be by having the tab itself slide out, as well as allow for multiple tabs.

At this point I'm happy enough to move on with a successful proof of concept, but I think if I were doing this from scratch again I wouldn't bother having the tab slide out as well.  I'd just show the sliding out window.

But, this should be a good starting point.

visualstudioflyoutmenus.htm (17.7 KB)

Update: And of course it completely fails in FF.

I made some changes, removed some things, tweaked others... looks ok in FF now.

visualstudioflyoutmenus2.htm (17.7 KB)

What the heck?  This ad does in fact go to Mozy.com.

 

You can use XXCOPY (www.xxcopy.com) to write simple backup scripts to copy your files to remote storage locations, but you can also use it to compress those files on the NTFS file system.

I have used /CLONE in the past because it's easy, but a side effect is that all the files will maintain their attributes when copied to the destination.  What this means is that if a file is not compressed at the source, it won't be compressed at the destination, even if you set the destination folder/drive to be compressed.

To solve this problem you just need to first set the compression attribute of the container of the location you are copying files too, and then use a switch to tell xxcopy to only set the "normal" attribute, which I belive sets the "A" attribute.  Either way, it works.

xxcopy E:\ /s /ASDC
xxcopy E:\ /s /AS:C
xXcopy D:\FolderRoot\*.* E:\FolderRoot /CLONE /YY /Z0 /KN 

The code here is backing up everything in D:\FolderRoot to the E: with compression.

The first step sets the E: to be a compressed drive (this will work for any folder, not just the root of the drive).

The second step sets all files in all folders to be compressed.  This is just to make sure that when this runs at night, no files are missed.

The last step copies all files from the folder we care about, and the /KN swtich allows the destination files to inherit the "C" (compression) attribute from the parent folder.

 

Because I didn't like the options for integrating SQL Server code into TFS, I wrote an application that scripts all our database objects to text files and then checks those files into TFS every night.  If there is a change in any of the files, those changes are versioned in TFS so we can go back and see when things were changed or recover old code without restoring the entire database.

But recently there was a problem with the server that my app runs on, and during the restore process, the backup process was run twice on a set of folders (at least that is my best guess) leaving thousands of duplciate files in a myriad of folders.

So for example, if there was a file called:

$\DatabaseFiles\sqlserver1\SomeDBName\SPs\dbo.SomeSP.sql

then the restore process created another file called

$\DatabaseFiles\sqlserver1\SomeDBName\SPs\Copy of dbo.SomeSP.sql

The problem is that my application does bulk checkins each night of all files under the \DatabaseFiles folder.  I didn't realize that the backup process had been messed up until it was too late, and those thousands of files were already checked into TFS.

So using this page to help with some of the syntax of using the TF.EXE command line too, I was able to perform a recursive wildcard delete.

In the end something like this worked just great:

tf.exe delete /recursive /noprompt "C:\TFSroot\DatabaseFiles\Copy of*"

Notice that the wildcard is applied to all recursive folders under C:\TFSroot\DatabaseFiles\.

Nice.

 

We ran into an interesting problem today.

A matrix report in SSRS (sqlserver reporting services) 2008 would work just fine when previewing it in VS, and would work fine when viewed directly on the reporting server.  But, if you view it through a asp.net reportviewer control, it would just show the header, a big blank space, and then the footer.

This only happens if you have Vertical text for the row headers.  Remove that and everything is OK.

I began editing the generated CSS/HTML and found that the cells had a number of styles applied, but specifically the one that seemed to break everything was:

WIDTH:100%;

Remove that and the page rendered as expected.

We tried changing a number of parameters to get it to remove the width style but no luck.

We have something that generates images with GDI to do it now, but it's not ideal.

NitroPDF.com offers a simplified version of their Pro online pdf product for free.

I needed to remove a few pages from a PDF document, and it worked perfect.

http://www.nitropdf.com/free/hammer/index.htm

Last night I had another instance where doing a:

SELECT * FROM
TBL WHERE ID NOT IN (SELECT ID FROM TBL2)

would fail every time to return any rows.

After longer than it should have been, I realized that it was this stupid issue that I blogged about already (SQL "NOT IN" Will Fail If The List Contains A Null).