Benchmark IT Consulting » General

What You Waiting for…….Worker Thread

Colin Stasiuk — Tue, 07 Sep 2010 19:46:54 +0000

The world needs more Gwen IMO

“Gwen Stefani – What you waiting for”

So I happened to have my handy Job Activity Monitor up as I was looking to see if a job I kicked off finished when I noticed a status of “Waiting for Worker Thread”. I can’t say that I’ve seen this message before so I started looking around as to what else was running at this time.

I had a couple PowerShell jobs running and this job that was “waiting for a worker thread” was a PowerShell job as well. I called my best friend Google up to see if there was something quick and easy out there to help me figure this one out and I got sent here:

http://support.microsoft.com/kb/306457

specifically…. “Note Microsoft SQL Server 2005 stores the max_worker_thread setting in the syssubsystems table in the msdb system database instead of in the registry. The syssubsystems table can be updated to reflect a non-default max_worker_thread setting.”

So I queried the syssubsystems table in msdb and guess what I found….

PowerShell by default has a max_worker_threads value of 2.

So a quick update to that value (in DEV… you didn’t actually think I would update a system table value in production did you?). I kicked off more than 2 jobs that called PowerShell scripts and BOOYEAH problem solved. Now I just bumped this up to 5 because I wanted to do some investigation as to any “gotchas” from messing with this value.

(so far I haven’t found any but I’ll update this post if any come up)

Enjoy!!

Oracle for the SQL DBA… Part 1 of a bunch LOL

Colin Stasiuk — Fri, 11 Jun 2010 17:07:50 +0000

So i’ve been a bit quiet lately as things have been very crazy in my world… which is good cause that means I’m staying busy and out of trouble but I thought I would drop a quick horribly long blog about some of the stuff I’m working on right now. One thing that seems to of landed on my plate is doing some Oracle DBA type stuff.

Now I am NOT an Oracle DBA and this post will prove that so if you are an Oracle DBA and have something easier then my final “solution” please drop me some comments cause this seems way to painful to be the “right” way LOL. As much as I’m an evangelist for SQL Server I’m wanting to learn more and more Oracle because you never know when something will get thrown over the fence right?

The following “use-case” came my way… “We’d like to have a fresh copy of our PROD database over to a new DEV database leaving the original DEV database intact”.

Seems easy enough… in my SQL Server environment I would restore from last nights backup of production and specify a new database name and filenames, paths, etc during the restore and bippidy-boppidy-bo we’re done like dinner.

This is where my journey of pain begins LOL (this is a long journey so feel free to skip to the end for my final solution or laugh at me along the way for your cruel fun and enjoyment hahaha)

I applied the logic I would use in SQL Server and thought about doing a “Database Recovery” using the Oracle Enterprise Manager 10G. Hmmmm in order to use Enterprise Manager you need to be logged into a database instance though… I can’t log into an instance because I’m trying to restore one… I can’t restore one cause I can’t log into an instance… I can’t log into an instance because I’m trying to restore one… I can’t restore one cause I can’t log into an instance… *sigh*

Next stop… Database Configuration Assistant. I thought I might find something useful there…. and I kind of did with “Create a Database”. So I created a general purpose database cause I thought it’ll just be a place holder for my restore anyways so next, next, next, next, finish. BOOYEAH! New database created. For the sake of this post we’ll call this MYDBDEV2.

OK back to Enterprise Manager… and I start down the road of recovery

(Guilty Pleasure = Eminem)

So I go to perform a Whole Database Recovery and Enterprise Manager gives me this…

“The database will be shut down and brought to the NOMOUNT (STARTED) state to restore the control file first. Are you sure you want to shut down the database now?”

Sure… that sounds good.

OK so back to “Perform a Whole Database Recovery”… Can I just take a moment to complain about the “save as preferred” checkbox that NEVER saves… I digress lol

OK so I go to restore “Directly from a specified backup” and enter in the path to my backup file and my next stop is: “The operation for restoring the controlfile may take some time to complete. Are you sure you want to execute it now?”

Sure… that sounds good. (All the Oracle folks reading this at this point I’m sure are laughing… but that’s fine LOL I can laugh now too about it)

Hmmmm so I viewed the details and it tells me… “ORA-01103: database name ‘MYDBPROD’ in control file is not ‘MYDBDEV2′”

*sigh* well yes I know that… so we’ll fast forward through the rest a bit here and give the quick run down of the rest of my pain before what I finally came up with that worked… so here we go:

Deleted MYDBDEV2 and created a new database named the same name as the production database MYDBPROD
That failed because I didn’t have the same file structure
I used a “database template” of MYDBPROD to create a database called MYDBPROD on the DEV server
I restored overtop of the MYDBPROD on the DEV server with a backup but that wouldn’t start cause my REDOLOGs were wrong
Deleted the REDOLOGs and it started up
Then the TEMP01.DBF file was giving me problems
Added a new TEMP.DBF file, set it to default, dropped the old TEMP01.DBF file
Went through a bunch of pain with NID command to rename the database
I mean a BUNCH of pain
MEGA, MEGA PAIN
Eventually got it to work

WOW…. ouch.

I then started playing around with the “Clone Database” tool in Enterprise Manager and here is what my final solution/steps look like: (This is where I want Oracle DBAs to comment on an easier/better solution LOL)

MY FINAL SOLUTION (for now LOL)

Create a new database (from a template of PROD) on the DEV server named the same as the database on the PROD server
Perform a recovery and restore the PROD backup over top of the DEV databases (that’s named the same as the PROD database)
Clone the DEV database that’s named the same as the PROD database and give it the name you want MYDBDEV2
Delete the DEV database that’s named the same as the PROD database

This CAN’T be the best way to do this… I just don’t believe it. So all you Oracle DBAs out there… you’ve had a good laugh at my expense LOL drop me some comments and help me help you (well I’m not actually helping you so that doesn’t really apply…but your help would be greatly appreciated :) )

In all honesty this was a really fun exercise for me. It’s nice to go into something totally blind and plug your way through it, get to a solution, find a better one, and keep on refining the process. I learned a tonne of stuff and am wanting to do and learn more and more! Oracle is like how PowerShell is/was for me… it’s fun to learn stuff that you don’t have a background in and after 13 years working with SQL Server I still learn new things but it’s a total different feeling with something so foreign to me.

Enjoy!!

EDMPASS July Meeting w/Brent Ozar #YEG

Colin Stasiuk — Mon, 31 May 2010 20:09:37 +0000

On July 7th 2010 the Edmonton Chapter of PASS is having it’s next meeting. Details below:

http://www.eventbrite.com/event/686883487

Please be sure to not only click the “Add to my calendar” but also the “Register” button so that we can plan accordingly for food and drinks.

Date: July 7th 2010
Time: 5:00 pm – 7:00 pm
Location: Stanley A. Milner library
Map: 7 Sir Winston Churchill Square
Meeting Room: Centennial Room – Basement
Speaker: Brent Ozar
Topic: Virtualization and Magicians: Dealing with Virtual SQL Server

Session Abstract:

Virtualization and Magicians: Dealing with Virtual SQL Server

So the sysadmins told you they’ve learned a few tricks, and they’re going to virtualize your database servers. In this session, we’ll talk about our real-world experiences with virtual SQL Servers and how to help the sysadmins avoid implementation mistakes that will saw SQL Server’s performance in half.

How to configure CPU, memory and storage for virtual environments
How to benchmark your servers and make sure they’ll handle your loads
Common pitfalls like antivirus definition downloads that can cripple your servers
What to watch for after V-day

Presenter Information: Brent Ozar

Brent is a SQL Server Domain Expert with Quest Software. Brent has a decade of broad IT experience, performing systems administration and project management before moving into database administration. In his current role, Brent specializes in performance tuning, disaster recovery and automating SQL Server management. Previously, Brent spent 2 years at Southern Wine & Spirits, a Miami-based wine & spirits distributor.

Brent has experience conducting training sessions, has written several technical articles, and blogs prolifically at http://www.BrentOzar.com. He’s currently writing SQL Server Internals and Troubleshooting for Wiley/Wrox along with Christian Bolton, and Justin Langford. He’s Editor-in-Chief at SQLServerPedia.com, where he also records video podcasts.

Agenda:
5:00 pm – Pizza and Socializing
5:30 pm – Sponsor Presentation
5:45 pm – Feature Presentation
7:00 pm – Wrap Up and Draws

If you haven’t signed up already at EDMPASS.com please do so now to receive meeting notifications, news, and updates from EDMPASS.

Hope to see you there.

Enjoy!!

List All Statistics For a Column

Colin Stasiuk — Tue, 11 May 2010 14:56:46 +0000

Msg 5074, Level 16, State 1, Line 1

The statistics ‘StatName’ is dependent on column ‘ColumnName’.

Msg 4922, Level 16, State 9, Line 1

ALTER TABLE DROP COLUMN ColumnName failed because one or more objects access this column.

So what is a nice quick and easy way to list all the statistics that are dependant on that column?

My googling skills are obviously subpar this morning cause I couldn’t find anything to help me and it seems to me that this would be something that lots of people have run into. This is what I came up with:

DECLARE @SearchColumnName	nvarchar(100)
SELECT	@SearchColumnName = 'SearchColumn'
 
SELECT	OBJECT_NAME(A.object_id) as 'TableName',
		C.stats_column_id AS 'StatisticID',
		B.name as 'StatisticName',
		COL_NAME(C.object_id, C.column_id) AS [Name]
FROM	sys.tables AS A	INNER JOIN
		sys.stats B ON B.object_id = A.object_id INNER JOIN
		sys.stats_columns C ON C.stats_id = B.stats_id AND C.object_id = B.object_id
WHERE	COL_NAME(C.object_id, C.column_id) = @SearchColumnName

I dunno… seems like there should be a sp_ListAllStatisticsForAColumn system stored proc or something LOL.

If you have something easier/better please post it in a comment!!

Enjoy!!

MAXRECURSION with a CTE in a View

Colin Stasiuk — Mon, 12 Apr 2010 14:37:10 +0000

So a minor “gotcha” here to start your day. If you’ve worked with CTEs before you’ve probably also worked with the MAXRECURSION option which specifies how many recursion levels deep to go before cancelling a statement.

(Code below borrowed from: http://msdn.microsoft.com/en-us/library/ms175972.aspx)

USE AdventureWorks;
GO
--Creates an infinite loop
WITH cte (EmployeeID, ManagerID, Title) as
(
    SELECT EmployeeID, ManagerID, Title
    FROM HumanResources.Employee
    WHERE ManagerID IS NOT NULL
  UNION ALL
    SELECT cte.EmployeeID, cte.ManagerID, cte.Title
    FROM cte
    JOIN  HumanResources.Employee AS e
        ON cte.ManagerID = e.EmployeeID
)
--Uses MAXRECURSION to limit the recursive levels to 2
SELECT EmployeeID, ManagerID, Title
FROM cte
OPTION (MAXRECURSION 2);
GO

Now if you wanted to create a view based on your CTE and SELECT statement and included your MAXRECURSION option you’d get an error message

Msg 156, Level 15, State 1, Procedure numbers, Line 15
Incorrect syntax near the keyword ‘OPTION’.

In order to make use of the MAXRECURSION option you need to first create your view without using the MAXRECURSION option:

USE AdventureWorks;
GO
CREATE VIEW vwCTE AS
--Creates an infinite loop
WITH cte (EmployeeID, ManagerID, Title) as
(
    SELECT EmployeeID, ManagerID, Title
    FROM HumanResources.Employee
    WHERE ManagerID IS NOT NULL
  UNION ALL
    SELECT cte.EmployeeID, cte.ManagerID, cte.Title
    FROM cte
    JOIN  HumanResources.Employee AS e
        ON cte.ManagerID = e.EmployeeID
)
-- Notice the MAXRECURSION option is removed
SELECT EmployeeID, ManagerID, Title
FROM cte
GO

Then when you query the view include the MAXRECURSION option:

USE AdventureWorks;
GO
SELECT 	EmployeeID, ManagerID, Title
FROM 	vwCTE
OPTION (MAXRECURSION 2);

Nothing too deep for this morning but whenever I have a “so THAT’s how it works” moment I like to share

Enjoy!!

SQL Server 2008 Internals and Troubleshooting and DBA Survivor

Colin Stasiuk — Tue, 16 Mar 2010 16:39:06 +0000

OK so I’m back from an extended vacation from BenchmarkITConsulting.com and although my vacation had nothing to do with warm weather, beaches, or bacon it was time well spent on finishing up my chapters for the soon to be released Pro Server 2008 Policy-Based Management book.

After going through this process I have a whole new respect for anyone who takes on the challenge of writing a book. It is a long and painful process that will eat up any and all free time that you have but it is probably one of the most rewarding challenges that I’ve taken on and I can’t wait to get my own “AUTHORS COPIES” box in the mail (great post/video BTW Tom)

2 books that are on my “must read” list now that I’m going to have more free time are:

SQL Server 2008 Internals and Troubleshooting and DBA Survivor

Brent and Tom both have always been great resources to help troubleshoot, bounce ideas off of, and just always willing to lend a hand in anything I’ve sent their way and I’m quite interested to give their books a read and a review. Tom is the technical editor for my book and although he never really explained what he meant by “punch harder” I think/hope I got the idea LOL.

I’ll admit that I haven’t picked up a book since starting this whole co-authoring a book started back in September/09. So I thought these two books would be a great place to start getting back into reading because the authors have both helped shape and guide my SQL path greatly over the last year or two.

So here’s to getting back into reading, blogging more often, and just generally getting back in touch with the world LOL

Enjoy!!

sp_refreshview How Did I Forget You?

Colin Stasiuk — Wed, 03 Mar 2010 20:37:50 +0000

So a developer asked me a question about how he added a column to a table and now the view that referenced the table wasn’t showing the column. I asked the obvious question: “Did you add the column to the view?” to which he replied “The view is doing a SELECT *”

OK it may not be that bad… and it definitely not the reason for this post as I think enough people have wrote articles on why you shouldn’t use SELECT * …. so anyways I queried the view myself and saw the same thing that the developer saw.

So after a little “googling” I came across sp_refreshview.

sp_refreshview (Transact-SQL)

“Updates the metadata for the specified non-schema-bound view. Persistent metadata for a view can become outdated because of changes to the underlying objects upon which the view depends.”

So I’m not sure if it’s because I don’t develop alot these days or if it’s because I name all my columns in my views directly but sp_refreshview slipped my mind so I thought I’d write a blog post about it just in case I’m not the only one.

Other ways to resolve this would be to do an ALTER VIEW and just keep the schema the same or to DROP and CREATE the view again (but you risk losing all permissions on the view so that’s just a bad idea). If Microsoft gives you a “sp_refreshview” just for this reason… I say use it!

EDIT: Denis Gobo tweeted me with a great article he wrote as well on the subject at LessThanDot

Enjoy!!

Rebooting a Server Using PowerShell

Colin Stasiuk — Mon, 01 Feb 2010 17:43:05 +0000

The more I work with PowerShell the more I love it

So what do you do if you’re in a VM environment and you want/need to reboot your server but you can’t connect via RDP and you can’t connect to the console using the VM Infrastructure Client, and all other attempts to connect just don’t work. I decided to go down the route of using PowerShell.

$ServerToRestart = Get-WMIObject win32_operatingsystem -computer <>
$ServerToRestart.reboot()

Get-WMIObject – Windows PowerShell: The WMI Connection

Simple, easy and very dangerous/powerful. I’ll channel my inner “Uncle Ben” (see Spiderman quotes) and just leave this message as a closing thought:

“With Great Power Comes Great Responsibility”

Enjoy!!

My 2010 Goals and Themeword

Colin Stasiuk — Wed, 30 Dec 2009 20:13:53 +0000

I’ve seen a few of these types of posts over the last week or so and I figured I should publish something on the intertubes that would be something to look back on come next year at this time. Tom LaRock had a very interesting spin on things (as he always does) and added a ThemeWord into the mix.

Colin’s 2010 Themeword is……… FOCUS

No not that kind of focus….

You don’t think of them as human… You don’t think of them at all

Colin Stasiuk — Tue, 24 Nov 2009 16:00:30 +0000

OK maybe I’m feeling really good today… maybe it’s because it’s @Peschkaj‘s birthday… maybe it’s because I needed to find a way to work some Tina Turner into my next blog post…or maybe it’s because I successfully took care of both kids last night for the first time while wifey started some photography classes.

Developers gets a bum rap… ok I have to stop right there… bum rap? who came up with that phrase?

http://www.yourdictionary.com/idioms/bum-rap

“This expression originated in the 1920s as underworld slang, and by the mid-1900s it was also used figuratively for other kinds of injustice.”

OK so apparently I’m going 1920′s with this blog post… back to developers

So yeah as DBAs we’re bred to hate on anyone (other then ourselves) who make changes to the environment that we’ve been charged with managing and protecting. We don’t trust developers and we think they are constantly trying to intentionally mess with “our world” and implement bad SQL code.

The relationship between DBA and developer is just like any other relationship in life… it requires work from both sides to be successful.

Ask yourself this:

Do you meet with the development team to get an idea of what projects they’re working on? their timelines? if they need any ideas or help, etc?
When was the last time you did a lunch and learn to try and help educate and show off some new tools or features in SQL Server?
When you “reject” or complain about developer code do you not only identify the problem but help provide alternative approaches or solutions?

Pssssssssssssssssssssst…Oh and by the way… as shocking as this sounds… come here… I’ve got a secret to tell you….

IT’S NOT YOUR DATABASE

Now I don’t want to get hung up on semantics but if the business didn’t require applications and developers didn’t create the applications and databases… what database environment would you be managing? LOL

At the end of the day we need all need to get along. Developers need to stop looking at DBAs as “road blocks” and DBAs need to stop looking at developers like terrorists who are just planning their next attack on your database environment.

So next time you have the urge to go rip a strip off an unsuspecting developer stop… breathe…. think about this blog post and come into the discussion not just armed with problems and complaints… but with solutions and education.

<< GROUP HUG >>

Enjoy!!