Thursday, August 31, 2006

sys.dm_exec_sessions

As I promised earlier today, here is the first of many posts about the Dynamic Management Views in SQL Server 2005.

What does the sys.dm_exec_sessions view return? It looks like it is a combination of DBCC USEROPTION and sp_who. Let's for example find out how many sessions are running, sleeping or dormant

Just a quick count of all the sessions
SELECT COUNT(*) as StatusCount,CASE status
WHEN 'Running' THEN 'Running - Currently running one or more requests'
WHEN 'Sleeping ' THEN 'Sleeping - Currently running no requests'
ELSE 'Dormant – Session is in prelogin state' END status
FROM sys.dm_exec_sessions
GROUP BY status

Let's list all SPID's
SELECT session_id,CASE status
WHEN 'Running' THEN 'Running - Currently running one or more requests'
WHEN 'Sleeping ' THEN 'Sleeping - Currently running no requests'
ELSE 'Dormant – Session is in prelogin state' END status
FROM sys.dm_exec_sessions


Let's just grab our own SPID
SELECT session_id,CASE status
WHEN 'Running' THEN 'Running - Currently running one or more requests'
WHEN 'Sleeping ' THEN 'Sleeping - Currently running no requests'
ELSE 'Dormant – Session is in prelogin state' END status
FROM sys.dm_exec_sessions
WHERE session_id = @@SPID


Now let's query all user sessions
SELECT * FROM sys.dm_exec_sessions
WHERE is_user_process =0

Then for the server you would change 0 to 1
SELECT * FROM sys.dm_exec_sessions
WHERE is_user_process =1

Just a quick count of all the transaction isolation levels
SELECT COUNT(*),CASE transaction_isolation_level
WHEN 0 THEN 'Unspecified'
WHEN 1 THEN 'ReadUncomitted'
WHEN 2 THEN 'Readcomitted'
WHEN 3 THEN 'Repeatable'
WHEN 4 THEN 'Serializable'
WHEN 5 THEN 'Snapshot' END AS TRANSACTION_ISOLATION_LEVEL
FROM sys.dm_exec_sessions
GROUP BY transaction_isolation_level



Let's look at some of these ANSI settings and while we are at it let's use UNPIVOT to return the results vertically
SELECT SPID,Value,ANSI_SETTING
FROM (
SELECT @@SPID as SPID,
CASE quoted_identifier
WHEN 1 THEN 'SET' ELSE 'OFF' END QUOTED_IDENTIFIER,
CASE arithabort
WHEN 1 THEN 'SET' ELSE 'OFF' END ARITHABORT,
CASE ansi_null_dflt_on
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_NULL_DFLT_ON,
CASE ansi_defaults
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_DEFAULTS ,
CASE ansi_warnings
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_WARNINGS,
CASE ansi_padding
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_PADDING,
CASE ansi_nulls
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_NULLS,
CASE concat_null_yields_null
WHEN 1 THEN 'SET' ELSE 'OFF' END CONCAT_NULL_YIELDS_NULL
FROM sys.dm_exec_sessions
WHERE session_id = @@SPID ) P
UNPIVOT (Value for ANSI_SETTING IN(
QUOTED_IDENTIFIER,ARITHABORT,ANSI_NULL_DFLT_ON,
ANSI_DEFAULTS,ANSI_WARNINGS,
ANSI_PADDING,ANSI_NULLS,CONCAT_NULL_YIELDS_NULL
)
) AS unpvt



Now let's see if we can duplicate DBCC USEROPTIONS
First run DBCC USEROPTIONS
And then run the following:

SELECT @@SPID as SPID,
CASE quoted_identifier
WHEN 1 THEN 'SET' ELSE 'OFF' END QUOTED_IDENTIFIER,
CASE arithabort
WHEN 1 THEN 'SET' ELSE 'OFF' END ARITHABORT,
CASE ansi_null_dflt_on
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_NULL_DFLT_ON,
CASE ansi_defaults
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_DEFAULTS ,
CASE ansi_warnings
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_WARNINGS,
CASE ansi_padding
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_PADDING,
CASE ansi_nulls
WHEN 1 THEN 'SET' ELSE 'OFF' END ANSI_NULLS,
CASE concat_null_yields_null
WHEN 1 THEN 'SET' ELSE 'OFF' END CONCAT_NULL_YIELDS_NULL,
CASE transaction_isolation_level
WHEN 0 THEN 'Unspecified'
WHEN 1 THEN 'ReadUncomitted'
WHEN 2 THEN 'Readcomitted'
WHEN 3 THEN 'Repeatable'
WHEN 4 THEN 'Serializable'
WHEN 5 THEN 'Snapshot' END AS TRANSACTION_ISOLATION_LEVEL,lock_timeout,date_first,date_format
FROM sys.dm_exec_sessions
WHERE session_id = @@SPID

It's probably not exactly the same and I don't know if I missed a column but below is the whole definition of the view and you can experiment with this yourself

Below is what the sys.dm_exec_sessions view returns
On the first line are the column name and the data type, below that is the description

sys.dm_exec_sessions

session_id smallint
Identifies the session associated with each active primary connection.

login_time datetime
Time when session was established.

host_name nvarchar(128)
Host associated with the session.

program_name nvarchar(128)
Program associated with the session.

host_process_id int
Process ID associated with the session.

client_version int
Version of the interface used by the client to connect to the Server.

client_interface_name nvarchar(32)
Name of the interface used by the client to connect to the Server.

security_id varbinary(85)
Microsoft Windows security ID associated with the login.

login_name nvarchar(128)
SQL Login Name associated with the session.

nt_domain nvarchar(128)
Domain from which session connection was made.

nt_user_name nvarchar(128)
Name of the user associated with the session.

status nvarchar(30)
Status of the session. Possible values:
-- Running - Currently running one or more requests
-- Sleeping - Currently running no requests
-- Dormant – Session is in prelogin state



context_info varbinary(128)
CONTEXT_INFO value for the session.

cpu_time int
CPU time, in milliseconds, that was used by this session.

memory_usage int
Number of 8-KB pages of memory used by this session.

total_scheduled_time int
Total time, in milliseconds, for which the session (requests within) were scheduled for execution.

total_elapsed_time int
Time, in milliseconds, since the session was established.

endpoint_id int
ID of the Endpoint associated with the session.

last_request_start_time datetime
Time at which the last request on the session began. This includes the currently executing request.

last_request_end_time datetime
Time of the last completion of a request on the session.

reads bigint
Number of reads performed, by requests in this session, during this session.

writes bigint
Number of writes performed, by requests in this session, during this session.

logical_reads bigint
Number of logical reads that have been performed on the session.

is_user_process bit
0 if the session is a system session. Otherwise, it is 1.

text_size int
TEXTSIZE setting for the session.

language nvarchar(128)
LANGUAGE setting for the session.

date_format nvarchar(3)
DATEFORMAT setting for the session.

date_first smallint
DATEFIRST setting for the session.

quoted_identifier bit
QUOTED_IDENTIFIER setting for the session.

arithabort bit
ARITHABORT setting for the session.

ansi_null_dflt_on bit
ANSI_NULL_DFLT_ON setting for the session.

ansi_defaults bit
ANSI_DEFAULTS setting for the session.

ansi_warnings bit
ANSI_WARNINGS setting for the session.

ansi_padding bit
ANSI_PADDING setting for the session.

ansi_nulls bit
ANSI_NULLS setting for the session.

concat_null_yields_null bit
CONCAT_NULL_YIELDS_NULL setting for the session.

transaction_isolation_level smallint
Transaction isolation level of the session.
-- 0 = Unspecified
-- 1 = ReadUncomitted
-- 2 = ReadCommitted
-- 3 = Repeatable
-- 4 = Serializable
-- 5 = Snapshot

lock_timeout int
LOCK_TIMEOUT setting for the session. The value is in milliseconds.

deadlock_priority int
DEADLOCK_PRIORITY setting for the session.

row_count bigint
number of rows returned on the session up to this point.

prev_error int
ID of the last error returned on the session.



So that's it for today, the view contains a lot more than I covered. i might expand this if I have time. If you have any suggestions leave me a comment or drop me an Email

The link to the post with all the views is here (as of today this is the only view that is covered)

Dynamic Management Views Blog Post Series Coming To A Screen Near You

Dynamic Management Views Blog Post Series Coming To A Screen Near You

Since I don't always know what to write I decided to give myself enough material for at least 6 months (if I cover 1 view a week). So here it is I will write about the Dynamic Management Views in SQL server 2005. The first view that I will cover is sys.dm_exec_sessions. You can use sys.dm_exec_sessions instead of DBCC USEROPTIONS only it's much better since you can query multiple SPID's

I will have the first post ready about 4:30 PM EST today

Here is a list of all the Dynamic Management Views in SQL Server 2005, I will link the ones that I write about to the post so that you can access all of this from 1 post (in theory)


Common Language Runtime Related Dynamic Management Views
sys.dm_clr_appdomains
sys.dm_clr_loaded_assemblies
sys.dm_clr_properties
sys.dm_clr_tasks

Database Mirroring Related Dynamic Management Views
sys.dm_db_mirroring_connections

Database Related Dynamic Management Views
sys.dm_db_file_space_usage
sys.dm_db_session_space_usage
sys.dm_db_partition_stats
sys.dm_db_task_space_usage

Execution Related Dynamic Management Views and Functions
sys.dm_exec_background_job_queue
sys.dm_exec_background_job_queue_stats
sys.dm_exec_cached_plans
sys.dm_exec_connections
sys.dm_exec_cursors
sys.dm_exec_plan_attributes
sys.dm_exec_query_optimizer_info
sys.dm_exec_query_plan
sys.dm_exec_query_stats
sys.dm_exec_requests
sys.dm_exec_sessions added on 2006-08-31
sys.dm_exec_sql_text

Full-Text Search Related Dynamic Management Views
sys.dm_fts_active_catalogs
sys.dm_fts_crawls
sys.dm_fts_crawl_ranges
sys.dm_fts_memory_buffers
sys.dm_fts_memory_pools

Index Related Dynamic Management Views and Functions
sys.dm_db_index_operational_stats
sys.dm_db_index_physical_stats
sys.dm_db_index_usage_stats added on 2006-09-10

I/O Related Dynamic Management Views and Functions
sys.dm_io_backup_tapes
sys.dm_io_cluster_shared_drives
sys.dm_io_pending_io_requests
sys.dm_io_virtual_file_stats

Query Notifications Related Dynamic Management Views
sys.dm_qn_subscriptions

Replication Related Dynamic Management Views
sys.dm_repl_articles
sys.dm_repl_schemas
sys.dm_repl_tranhash
sys.dm_repl_traninfo

Service Broker Related Dynamic Management Views
sys.dm_broker_activated_tasks
sys.dm_broker_connections
sys.dm_broker_forwarded_messages
sys.dm_broker_queue_monitors

SQL Operating System Related Dynamic Management Views
sys.dm_os_buffer_descriptors
sys.dm_os_memory_pools
sys.dm_os_child_instances
sys.dm_os_performance_counters
sys.dm_os_cluster_nodes
sys.dm_os_schedulers
sys.dm_os_hosts
sys.dm_os_stacks
sys.dm_os_latch_stats
sys.dm_os_sys_info
sys.dm_os_loaded_modules
sys.dm_os_tasks
sys.dm_os_memory_cache_clock_hands
sys.dm_os_threads
sys.dm_os_memory_cache_counters
sys.dm_os_virtual_address_dump
sys.dm_os_memory_cache_entries
sys.dm_os_wait_stats
sys.dm_os_memory_cache_hash_tables
sys.dm_os_waiting_tasks
sys.dm_os_memory_clerks
sys.dm_os_workers
sys.dm_os_memory_objects

Transaction Related Dynamic Management Views and Functions
sys.dm_tran_active_snapshot_database_transactions
sys.dm_tran_active_transactions
sys.dm_tran_current_snapshot
sys.dm_tran_current_transaction
sys.dm_tran_database_transactions
sys.dm_tran_locks
sys.dm_tran_session_transactions
sys.dm_tran_top_version_generators
sys.dm_tran_transactions_snapshot
sys.dm_tran_version_store

Wednesday, August 30, 2006

A SQL Super Hero Is Born

Ken Henderson promised to give away a signed book to the person who made the best Celko superhero
Tom Øyvind Hogstad has created one on his blog and I think it looks pretty good, even the S in the logo can stay; we can assume it stands for SQL




Below you will find my favorite Celko answer ever

>> how to set the PK in SS Mgmt Studio ? <<

Who cares?? You are not not a real SQL programmer!! You are a "mousey, mousey, click , click" non-programmer. (with a French accent) we spit on you, Video gamer! to be serious, real programmers use a text editor. They know the language they write in. Those stinking "video game tools"slow us down. And they lead us to ask questiosn like this in newsgroups where people liek me will maek fun of you.

You can find more of these Celko ‘answers on Joe Celko The SQL Apprentice

Tuesday, August 29, 2006

Visual Studio 2005 Team Edition For Database Professionals Review On Regdeveloper

The Register has a review on their subsite Regdeveloper about Visual Studio 2005 Team Edition For Database Professionals (also know as Data Dude) There are 3 pictures so that you can see what the tool looks like if you are afraid to install the latest Community Technology Preview on your system( you can get the latest release here )

The review is very positive about the tool and this will be one of those tools that you have to have in your shop, just as is the case with SQL Compare and SQL LiteSpeed right now.

There is one little problem with the review in the following sentence: "DD was partially the brainchild of Ewan Garden and Gert Drapers" they managed to spell Euan Garden's name wrong

Read the review here

Monday, August 28, 2006

Round Up Or Down To Nearest Percentage Value By Using FLOOR And CEILING

Let's say you have a value of 13.33 and you want to round this up and down to the nearest .25 in other words for 13.33 you want to display 13.25 and 13.50.How do you do that? It's pretty easy you do FLOOR(Value *4)/4 and CEILING(Value *4)/4
Below are 3 example, 1 for 0.25, 1 for 0.50 and 1 for 0.33


--0.25
DECLARE @Value DECIMAL(10,2)

SET @Value = 13.33

SELECT FLOOR(@Value * 4) / 4.0,CEILING(@Value * 4) / 4.0
GO


--0.50
DECLARE @Value DECIMAL(10,2)

SET @Value = 13.33

SELECT FLOOR(@Value * 2) / 2.0,CEILING(@Value * 2) / 2.0
GO


--0.33
DECLARE @Value DECIMAL(10,2)

SET @Value = 13.36

SELECT FLOOR(@Value * 3) / 3.0,CEILING(@Value * 3) / 3.

Friday, August 25, 2006

Win One Of Ken Henderson's SQL Server Guru Books By Creating A PhotoShop Joe Celko Action Figure

That's right you can win one of the following three books

The Guru's Guide to Transact-SQL
The Guru's Guide to SQL Server Architecture and Internals
The Guru's Guide to SQL Server Stored Procedures, XML, and HTML

All you have to do is create a photoshopped depiction of Joe Celko as an action figure.

The best one gets a signed copy of one of his books (too bad I already have all three)

Ken Henderson's Blog

Friday Joke and Frappr Map Update

Since it's Friday today I will keep it light
Let's start with a joke that I read in Wired magazine yesterday

What do you call a blonde who dyes her hair brown?

Artificial intelligence

Next topic

Yesterday I created a Frappr map and asked for you (the reader) to add yourself to the map so that I can see where some of you are located
As of now the countries that are marked are

United States 6 people
Russia 2 people
Canada 1 person
Norway 1 person

So not quite close to my IQ yet

Thursday, August 24, 2006

SqlServerCode Frappr Map

I created a Frappr Map and would like for you (yes you the reader) to add yourself to the map because it would be interesting to see where some of the readers of this blog are located
So far I have less members (2) than my IQ, I would like to have more members than my IQ (believe me not a hard task at all)

So this is your chance; beat my IQ

My member name is Denis The SQL Menace

Quest Toad for SQL Server 2.0 Released

Quest Software, Inc. announced the newest version of its award-winning Toad™ database management product family. Toad for SQL Server 2.0 includes new features such as one-of-a-kind integrated debugging, SQL optimization, and integrated comparison and synchronization capabilities for SQL Server 2005. These new features further improve the productivity of SQL Server developers and database administrators (DBAs).


New features in version 2.0 of Toad for SQL Server include:

  • Database Back-Up Indicator – Demonstrates when it is necessary to back up a system, and integrates seamlessly with LiteSpeed for SQL Server

  • SQL Optimization - Automates the process of identifying problematic SQL then re-writing and benchmarking it for optimal SQL Server performance.

  • Compare and Sync – Provides advanced integrated comparison and synchronization of schema, data and server settings of two SQL Server environments.

  • Debugging – Offers integrated debugging capabilities for SQL Server 2005.
“With Toad for SQL Server 2.0, Quest offers customers a single interface to develop, administer and optimize their SQL Server environments,” said Douglas Chrystall, chief architect, Quest Software. “In addition to this release, Quest is making its Comparison Suite of tools free of charge. This further demonstrates Quest’s commitment to the SQL Server market – not only by expanding the breadth and depth our SQL Server portfolio with the enhancements to Toad, but also making essential tools, such as the Comparison Suite, freely available to the market.”

Pricing and Availability
A freeware version of 2.0 of Quest Toad for SQL Server is available now at http://www.toadsoft.com/toadsqlserver/toad_sqlserver.htm

A full production version is also available now with North American pricing beginning at $470 per seat (USD). For more information and to download free evaluation copies please visit: http://www.quest.com/toad_for_sql_server/

Tuesday, August 22, 2006

DDL Trigger Events Revisited

Many many moons ago I wrote about DDL Trigger Events
I promised that I would find out all the events that you can combine for example DDL_VIEW_EVENTS instead of CREATE_VIEW
,ALTER_VIEW and DROP_VIEW
So I finally did it, I created a bunch of triggers and then checked with a query that joined sys.triggers and sys.trigger_events

I started by creating a trigger and I used DDL_VIEW_EVENTS

CREATE TRIGGER ddlTestEvents
ON DATABASE
FOR
DDL_VIEW_EVENTS
AS
PRINT
'You must disable Trigger "ddlTestEvents" to drop, create or alter Views!'
ROLLBACK;
GO

After that I would check the sys.triggers and sys.trigger_events views to see what was inserted

SELECT name,te.type,te.type_desc
FROM sys.triggers t
JOIN sys.trigger_events te on t.object_id = te.object_id
WHERE t.parent_class=0
AND name IN('ddlTestEvents')
ORDER BY te.type,te.type_desc

In this case 3 rows were inserted

DDL_VIEW_EVENTS
----------------------------------
41 CREATE_VIEW
42 ALTER_VIEW
43 DROP_VIEW

Then I altered the trigger as follows

CREATE TRIGGER ddlTestEvents
ON DATABASE
FOR
DDL_USER_EVENTS
AS
PRINT
'You must disable Trigger "ddlTestEvents" to drop, create or alter Users!'
ROLLBACK;
GO

I checked again with the same query

SELECT name,te.type,te.type_desc
FROM sys.triggers t
JOIN sys.trigger_events te on t.object_id = te.object_id
WHERE t.parent_class=0
AND name IN('ddlTestEvents')
ORDER BY te.type,te.type_desc
in this case 3 rows were also inserted


DDL_USER_EVENTS
----------------------------------
131 CREATE_USER
132 ALTER_USER
133 DROP_USER

I kept repeating this until I had all the events and below is the result
The number that you see before the event is the type from the sys.trigger_events view

DDL_XML_SCHEMA_COLLECTION_EVENTS
----------------------------------
177 CREATE_XML_SCHEMA_COLLECTION
178 ALTER_XML_SCHEMA_COLLECTION
179 DROP_XML_SCHEMA_COLLECTION


DDL_VIEW_EVENTS
----------------------------------
41 CREATE_VIEW
42 ALTER_VIEW
43 DROP_VIEW


DDL_TRIGGER_EVENTS
----------------------------------
71 CREATE_TRIGGER
72 ALTER_TRIGGER
73 DROP_TRIGGER


DDL_USER_EVENTS
----------------------------------
131 CREATE_USER
132 ALTER_USER
133 DROP_USER


DDL_TYPE_EVENTS
----------------------------------
91 CREATE_TYPE
93 DROP_TYPE


DDL_TABLE_EVENTS
----------------------------------
21 CREATE_TABLE
22 ALTER_TABLE
23 DROP_TABLE


DDL_SYNONYM_EVENTS
----------------------------------
34 CREATE_SYNONYM
36 DROP_SYNONYM


DDL_STATISTICS_EVENTS
-----------------------------------
27 CREATE_STATISTICS
28 UPDATE_STATISTICS
29 DROP_STATISTICS


DDL_SERVICE_EVENTS
------------------------------------
161 CREATE_SERVICE
162 ALTER_SERVICE
163 DROP_SERVICE


DDL_SCHEMA_EVENTS
------------------------------------
141 CREATE_SCHEMA
142 ALTER_SCHEMA
143 DROP_SCHEMA


DDL_ROUTE_EVENTS
------------------------------------
164 CREATE_ROUTE
165 ALTER_ROUTE
166 DROP_ROUTE


DDL_ROLE_EVENTS
----------------------------------
134 CREATE_ROLE
135 ALTER_ROLE
136 DROP_ROLE


DDL_REMOTE_SERVICE_BINDING_EVENTS
-----------------------------------
174 CREATE_REMOTE_SERVICE_BINDING
175 ALTER_REMOTE_SERVICE_BINDING
176 DROP_REMOTE_SERVICE_BINDING


DDL_QUEUE_EVENTS
---------------------------------
157 CREATE_QUEUE
158 ALTER_QUEUE
159 DROP_QUEUE


DDL_PROCEDURE_EVENTS
----------------------------------
51 CREATE_PROCEDURE
52 ALTER_PROCEDURE
53 DROP_PROCEDURE


DDL_PARTITION_SCHEME_EVENTS
------------------------------------
194 CREATE_PARTITION_SCHEME
195 ALTER_PARTITION_SCHEME
196 DROP_PARTITION_SCHEME


DDL_PARTITION_FUNCTION_EVENTS
------------------------------
191 CREATE_PARTITION_FUNCTION
192 ALTER_PARTITION_FUNCTION
193 DROP_PARTITION_FUNCTION


DDL_EVENT_NOTIFICATION_EVENTS
----------------------------------
74 CREATE_EVENT_NOTIFICATION
76 DROP_EVENT_NOTIFICATION


DDL_ASSEMBLY_EVENTS
-----------------------------------
101 CREATE_ASSEMBLY
102 ALTER_ASSEMBLY
103 DROP_ASSEMBLY


DDL_CONTRACT_EVENTS
-----------------------------------
154 CREATE_CONTRACT
156 DROP_CONTRACT


DDL_FUNCTION_EVENTS
---------------------------------
61 CREATE_FUNCTION
62 ALTER_FUNCTION
63 DROP_FUNCTION


DDL_INDEX_EVENTS
---------------------------------
24 CREATE_INDEX
25 ALTER_INDEX
26 DROP_INDEX
206 CREATE_XML_INDEX


DDL_MESSAGE_TYPE_EVENTS
------------------------------------
151 CREATE_MESSAGE_TYPE
152 ALTER_MESSAGE_TYPE
153 DROP_MESSAGE_TYPE

Thursday, August 17, 2006

Architecting Microsoft.com – SQL Server Podcast And Video

Channel 9 has an interesting video and podcast about the microsoft.com architecture with SQL Server
From the site: "At the heart of most web sites once you dig down below the routers, below the web servers, below the ASP.NET code you find a critical piece. The heart of the matter if you will is the database and if that heart stops... well let’s just say your website will assume room temperature quickly. The question is; what do the guys at Microsoft.com do to insure that the heart of the system is thumpin’ and pumpin’ day in and day out 24x7 while still needing to do updates, patches, server maintenance etc.? Well never fear because we have the answer for you so listen up and learn something!"

The podcast is available in WMA and MP3 formats and runs for 27 minutes and 55 seconds
The video length is 26 minutes and 9 seconds

you can get the podcast and the video here (http://channel9.msdn.com/Showpost.aspx?postid=226519)

Wednesday, August 16, 2006

Query Multiple Databases

"Query Multiple Databases"
This search term has been used 13 times in the last week on this blog. So here is a little blog post about it
It's really not complicated to query multiple databases. You have to have permissions on both databases
The syntax looks like this:
SELECT *
FROM Database1.ObjectOwner.TableName
JOIN Database2.ObjectOwner.TableName ON.....

So here is a little script so that you can test it out

USE master
GO

--Create the first Database
CREATE DATABASE DBtest1
GO

--Create the second Database
CREATE DATABASE DBtest2
GO



USE DBtest1
GO

--Create Table1 on DBtest1
CREATE TABLE Table1(id INT, DescriptionValue VARCHAR(49))
INSERT Table1 VALUES (1,'DBtest1')
GO


USE DBtest2
GO

--Create Table2 on DBtest2
CREATE TABLE Table2(id INT, DescriptionValue VARCHAR(49))
INSERT Table2 VALUES (1,'DBtest2')
GO



USE master
GO

--Do the join between DBtest1 and DBtest2
SELECT *
FROM DBtest1.dbo.Table1 t1
JOIN DBtest2.dbo.Table2 t2 ON t1.id = t2.id
GO


--Another way
SELECT *
FROM DBtest1..Table1 t1
JOIN DBtest2..Table2 t2 ON t1.id = t2.id
GO



--Drop these databases
DROP DATABASE DBtest1,DBtest2

Monday, August 14, 2006

Analysis Services Stored Procedure Project Released On CodePlex

The Analysis Services Stored Procedure project is a set of sample stored procedures for Analysis Services 2005. These samples have been written in C# and sample MDX queries are included that demonstrate the use of the procedures against the Adventure Works DW sample database. It was developed by a group of community volunteers with 2 main aims in mind:

To provide a set of useful extensions to Analysis Services 2005.
To provide a variety of example source code for people looking to write their own stored procedures.

Visit the CodePlex site to download the source code

Friday, August 11, 2006

Use The Sieve of Eratosthenes To Find All PrimeNumbers Below 1 Million

In mathematics, the Sieve of Eratosthenes is a simple, ancient algorithm for finding all prime numbers up to a specified integer. It was created by Eratosthenes, an ancient Greek mathematician. Wheel factorization is often applied on the list of integers to be checked for primality, before Sieve of Eratosthenes is used, to increase the speed.


Algorithm
Write a list of numbers from 2 to the largest number you want to test for primality. Call this List A. (This is the list of squares on the left-hand-side of the picture.)
Write the number 2, the first prime number, in another list for primes found. Call this List B. (This is the list on the right-hand-side of the picture.)
Strike off 2 and all multiples of 2 from List A.
The first remaining number in the list is a prime number. Write this number into List B.
Strike off this number and all multiples of this number from List A. The crossing-off of multiples can be started at the square of the number, as lower multiples have already been crossed out in previous steps.
Repeat steps 4 through 6 until no more numbers are left in List A.

Read more on Wikipedia


So here is the SQL Version, it runs in 7 seconds on my machine when I run it a
second time, first run is 16 seconds


SET NOCOUNT ON


DECLARE @i INT

-- Create a 10-digit table
DECLARE @D TABLE (N INT)
INSERT INTO @D (N)
SELECT 0 UNION ALL
SELECT 1 UNION ALL
SELECT 2 UNION ALL
SELECT 3 UNION ALL
SELECT 4

INSERT INTO @D (N)
SELECT N+5 FROM @D

-- build a small sieve table between 2 and 1000
DECLARE @T TABLE (N INT)
INSERT INTO @T( N )
SELECT 1+A.N+10*(B.N+10*C.N)
FROM @D A, @D B, @D C

DELETE FROM @T WHERE N = 1

SET @I = 2
WHILE @I <= SQRT(1000)
BEGIN
DELETE FROM @T WHERE N % @I = 0 AND N > @I
SET @I = @I + 1
END

-- Create large table between 1001 and 1000000
SELECT A+10*(B+10*(C+10*(D+10*(E+ 10*F)))) AS N
INTO #P
FROM
( SELECT A.N AS A, B.N AS B, C.N AS C, D.N AS D, E.N AS E, F.N AS F
FROM @D A, @D B, @D C, @D D, @D E, @D F
WHERE A.N in (1, 3, 7, 9) -- Not divisible by 2 or 5
) blah
WHERE (A+B+C+D+E+F) % 3 <> 0 -- Or 3
AND (A+3*B+2*C-D-3*E-2*F) % 7 <> 0 -- Or 7
AND (B-A+D-C+F-E) % 11 <> 0 -- Or 11
AND D|E|F <> 0 -- Don't include the first 1000 numbers,
--we already have these in the small sieve table
UNION ALL SELECT 1000000

-- sieve the big table with smaller one
SELECT @I = 2
WHILE @I IS NOT NULL
BEGIN
DELETE FROM #P WHERE N% @I = 0
SELECT @I = MIN(N) FROM @T WHERE N > @I
END

-- add primes up to 1000
INSERT INTO #P SELECT N FROM @T

-- Here are the results
--78498 rows
SELECT * FROM #P ORDER BY 1

drop table #P
go

Wednesday, August 09, 2006

Try Quest LiteSpeed And Win An iPod Nano (Hopefully)

Quest has a nice little give away. They are giving away 40 2GB Apple Nano iPods to people who are willing to try LiteSpeed for SQL Server.
I am already using LiteSpeed on the production machines but the cool thing is that Quest has a free version that you can use with the developer editions of SQL Server at no cost at all
So check it out here

Monday, August 07, 2006

SQL Server Book Sales Up 98% Compared To Last Year

Tim O’Reilly has posted his latest State of the Computer Book Market article. Here are 2 excerpts: “Looking at the Database rollup, we again see the strength of SQL Server, the decline of Oracle book sales, and that while MySQL is still a much larger category than Postgres, Postgres is showing some curious strength. This is one of the things that treemap visualizations are good for. Small, bright green categories stand out, and you can start paying closer attention. (Ruby also showed bright green while it was still a tiny category before its remarkable surge over the past year.) We also see the continuing popularity of personal databases like Access and Filemaker.



“A few high-level observations:
• C# book sales continues to gain on Java, with a 49% unit sales increase compared to Java's 10% decrease.
• Ruby continues its momentum, and is the fastest-growing programming language in terms of book sales.
• Microsoft's new release of SQL Server has continued to drive significant book sales, with that market up 86%. ASP.Net is also on a roll, with book sales up 61%.

Read the whole article here

Thursday, August 03, 2006

Store The Output Of A Stored Procedure In A Table Without Creating A Table

I saw this technique in the Inside Microsoft SQL Server 2005: T-SQL Querying T-SQL and today I saw it also here The OPENROWSET Trick: Accessing Stored Procedure Output In A SELECT Statement @ Ward Pond's SQL Server blog

I decided to expand this a little
Sometimes I need to quickly find out who is blocking what (or what is blocking who for that matter) so I run sp_who2 then look at the BlkBy column and run a DBCC INPUTBUFFER or fn_get_sql to get the sql statement

When you have a lot of connections it's a pain in the neck to look for the BlkBy where it's not . since the result is ordered by SPID
You can always do
CREATE TABLE
INSERT TABLE
EXEC Proc

But who wants to create tables all the time (not me, at least not for this stuff)
SPID is twice in the resultset of sp_who2 (who know why?) so that complicates things a little
So let's start with sp_who


SELECT * INTO #TempSpWho
FROM OPENROWSET ('SQLOLEDB','Server=(local);TRUSTED_CONNECTION=YES;','set fmtonly off exec master.dbo.sp_who')

--return only the connections to master
SELECT * FROM #TempSpWho
WHERE dbname ='master'



--Let's try sp_who2
SELECT * INTO #TempSpWho2
FROM OPENROWSET ('SQLOLEDB','Server=(local);TRUSTED_CONNECTION=YES;','set fmtonly off exec master.dbo.sp_who2')

--Oops
Server: Msg 2705, Level 16, State 3, Line 1
Column names in each table must be unique. Column name 'SPID' in table '#TempSpWho2' is specified more than once.

--No problem list the columns
SELECT SPID, Status, Login, HostName, BlkBy,DBName, Command, CPUTime, DiskIO, LastBatch, ProgramName
INTO #TempSpWho2
FROM OPENROWSET ('SQLOLEDB','Server=(local);TRUSTED_CONNECTION=YES;','set fmtonly off exec master.dbo.sp_who2')

--Get the connections to master only
SELECT * FROM #TempSpWho2
WHERE dbname ='master'

--Get the blocking SPID's
SELECT * FROM #TempSpWho2
WHERE BlkBy NOT LIKE '% .'

--Get user connections only (everything with a SPID below 51 is reserved for sql server processes)
SELECT * FROM #TempSpWho2
WHERE SPID >= 50

A couple of notes
The OPENROWSET call opens a separate connection to SQL Server, so there is some overhead associated with this approach
This technique is not supported inside a declared transaction
Blocking can occure with a poorly architected stored procedure

Wednesday, August 02, 2006

Top SQL Server Google Searches For July 2006

These are the top SQL Searches on this site for the month of July. I have left out searches that have nothing to do with SQL Server or programming. As you can see it has been a slow(er) month. And Google reports only searches that have been searched for more than once. For example if you have these 4 searches
insert triggers
insert trigger
insert triggers
insert trigger!

Then only insert triggers will show up since the other two were used only one time


Here are the results...

increase maximum row size in mssql
crystal report division by zero code
insert triggers
select sum where
select from
delete a null record


Let's talk about the last search (delete a null record ) a little bit
The code below will show you how exactly you delete a row with a NULL value. You can not use =NULL but you have to use IS NULL

CREATE TABLE #TestDeleteNull (id INT identity, SomeDate DATETIME)
INSERT #TestDeleteNull VALUES(GETDATE())
INSERT #TestDeleteNull VALUES(GETDATE())
INSERT #TestDeleteNull VALUES(NULL)


--Check what's in the table
SELECT * FROM #TestDeleteNull

--This won't work
DELETE #TestDeleteNull
WHERE SomeDate = NULL
--(0 row(s) affected)

--This is the way
DELETE #TestDeleteNull
WHERE SomeDate IS NULL
--(1 row(s) affected)

--Check again
SELECT * FROM #TestDeleteNull

Tuesday, August 01, 2006

6 Different Ways To Get The Current Identity Value From A Table

This article will show you how to get the current identity value from a table and also some things that might act a little different than you would expect

Let's first create our two simple tables
CREATE TABLE TestOne (id INT identity,SomeDate DATETIME)
CREATE TABLE TestTwo (id INT identity,TestOneID INT,SomeDate DATETIME)

--Let's insert 4 rows into the table
INSERT TestOne VALUES(GETDATE())
INSERT TestOne VALUES(GETDATE())
INSERT TestOne VALUES(GETDATE())
INSERT TestOne VALUES(GETDATE())



Here are 6 ways to check for the current value

--1 @@IDENTITY
SELECT @@IDENTITY
--this returns 4


--2 DBCC CHECKIDENT
DBCC CHECKIDENT (TestOne, NORESEED)
after running DBCC CHECKIDENT the message returned is
Checking identity information: current identity value '4', current column value '4'.
DBCC execution completed. If DBCC printed error messages, contact your system administrator.


--3 MAX function
SELECT MAX(id)
FROM TestOne
you can also check with the MAX function but this is not recommended becuase you might get some other identity value that is not yours but from a different user

--4 TOP 1 and ORDER BY DESC
SELECT TOP 1 id
FROM TestOne
ORDER BY id DESC
--The same applies here as for the max function, this is not recommended


--5 IDENT_CURRENT
SELECT IDENT_CURRENT('TestOne')
--IDENT_CURRENT is another way to check


--6 SCOPE_IDENTITY
SELECT SCOPE_IDENTITY()
--This one is very similar to @@IDENTITY with one BIG difference (shown later)



--Let's add a trigger to the TestOne table
CREATE TRIGGER trTestOne ON [dbo].[TestOne]
FOR INSERT
AS
DECLARE
@CreditUserID INT

SELECT @CreditUserID = (SELECT ID FROM Inserted)

INSERT TestTwo VALUES(@CreditUserID,GETDATE())
GO

--Let's insert another row into the TestOne table
INSERT TestOne VALUES(GETDATE())

SELECT @@IDENTITY --1
SELECT @@IDENTITY,SCOPE_IDENTITY(),IDENT_CURRENT('TestOne') --1,5,5

Now why doesn't @@IDENTITY return 5 but 1?
This is because @@IDENTITY doesn't care about scope and returns the last identity value from all the statements, which in this case is from the code within the trigger trTestOne




So the bottom line is this: Always use SCOPE_IDENTITY() unless you DO need the last identity value regradless of scope (for example you need to know the identity from the table insert inside the trigger)

--Clean up this mess
DROP TABLE TestOne,TestTwo