Search Results

Search found 62701 results on 2509 pages for 'sql function'.

Page 13/2509 | < Previous Page | 9 10 11 12 13 14 15 16 17 18 19 20  | Next Page >

  • SQL SERVER – Understanding ALTER INDEX ALL REBUILD with Disabled Clustered Index

    - by pinaldave
    This blog is in response to the ongoing communication with the reader who had earlier asked the question of SQL SERVER – Disable Clustered Index and Data Insert. The same reader has asked me the difference between ALTER INDEX ALL REBUILD and ALTER INDEX REBUILD along with disabled clustered index. Instead of writing a big theory, we will go over the demo right away. Here are the steps that we intend to follow. 1) Create Clustered and Nonclustered Index 2) Disable Clustered and Nonclustered Index 3) Enable – a) All Indexes, b) Clustered Index USE tempdb GO -- Drop Table if Exists IF EXISTS (SELECT * FROM sys.objects WHERE OBJECT_ID = OBJECT_ID(N'[dbo].[TableName]') AND type IN (N'U')) DROP TABLE [dbo].[TableName] GO -- Create Table CREATE TABLE [dbo].[TableName]( [ID] [int] NOT NULL, [FirstCol] [varchar](50) NULL ) GO -- Create Clustered Index ALTER TABLE [TableName] ADD CONSTRAINT [PK_TableName] PRIMARY KEY CLUSTERED ([ID] ASC) GO -- Create Nonclustered Index CREATE UNIQUE NONCLUSTERED INDEX [IX_NonClustered_TableName] ON [dbo].[TableName] ([FirstCol] ASC) GO -- Check that all the indexes are enabled SELECT OBJECT_NAME(OBJECT_ID), Name, type_desc, is_disabled FROM sys.indexes WHERE OBJECT_NAME(OBJECT_ID) = 'TableName' GO Now let us disable both the indexes. -- Disable Indexes -- Disable Nonclustered Index ALTER INDEX [IX_NonClustered_TableName] ON [dbo].[TableName] DISABLE GO -- Disable Clustered Index ALTER INDEX [PK_TableName] ON [dbo].[TableName] DISABLE GO -- Check that all the indexes are disabled SELECT OBJECT_NAME(OBJECT_ID), Name, type_desc, is_disabled FROM sys.indexes WHERE OBJECT_NAME(OBJECT_ID) = 'TableName' GO Next, let us rebuild all the indexes and see the output. -- Test 1: ALTER INDEX ALL REBUILD -- Rebuliding should work fine ALTER INDEX ALL ON [dbo].[TableName] REBUILD GO -- Check that all the indexes are enabled SELECT OBJECT_NAME(OBJECT_ID), Name, type_desc, is_disabled FROM sys.indexes WHERE OBJECT_NAME(OBJECT_ID) = 'TableName' GO Now, once again disable indexes for the second test. -- Disable Indexes -- Disable Nonclustered Index ALTER INDEX [IX_NonClustered_TableName] ON [dbo].[TableName] DISABLE GO -- Disable Clustered Index ALTER INDEX [PK_TableName] ON [dbo].[TableName] DISABLE GO -- Check that all the indexes are disabled SELECT OBJECT_NAME(OBJECT_ID), Name, type_desc, is_disabled FROM sys.indexes WHERE OBJECT_NAME(OBJECT_ID) = 'TableName' GO Next, let us build only the clustered index and see the output of all the indexes. -- Test 2: ALTER INDEX REBUILD -- Rebuliding should work fine ALTER INDEX [PK_TableName] ON [dbo].[TableName] REBUILD GO -- Check that only clustered index is enabled SELECT OBJECT_NAME(OBJECT_ID), Name, type_desc, is_disabled FROM sys.indexes WHERE OBJECT_NAME(OBJECT_ID) = 'TableName' GO Let us do final clean up. -- Clean up DROP TABLE [TableName] GO From the example, it is very clear that if you have built only clustered index when the nonclustered index is disabled, it still remains disabled. Do let me know if the idea is clear. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Index, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL SERVER – Solution – Puzzle – SELECT * vs SELECT COUNT(*)

    - by pinaldave
    Earlier I have published Puzzle Why SELECT * throws an error but SELECT COUNT(*) does not. This question have received many interesting comments. Let us go over few of the answers, which are valid. Before I start the same, let me acknowledge Rob Farley who has not only answered correctly very first but also started interesting conversation in the same thread. The usual question will be what is the right answer. I would like to point to official Microsoft Connect Items which discusses the same. RGarvao https://connect.microsoft.com/SQLServer/feedback/details/671475/select-test-where-exists-select tiberiu utan http://connect.microsoft.com/SQLServer/feedback/details/338532/count-returns-a-value-1 Rob Farley count(*) is about counting rows, not a particular column. It doesn’t even look to see what columns are available, it’ll just count the rows, which in the case of a missing FROM clause, is 1. “select *” is designed to return columns, and therefore barfs if there are none available. Even more odd is this one: select ‘blah’ where exists (select *) You might be surprised at the results… Koushik The engine performs a “Constant scan” for Count(*) where as in the case of “SELECT *” the engine is trying to perform either Index/Cluster/Table scans. amikolaj When you query ‘select * from sometable’, SQL replaces * with the current schema of that table. With out a source for the schema, SQL throws an error. so when you query ‘select count(*)’, you are counting the one row. * is just a constant to SQL here. Check out the execution plan. Like the description states – ‘Scan an internal table of constants.’ You could do ‘select COUNT(‘my name is adam and this is my answer’)’ and get the same answer. Netra Acharya SELECT * Here, * represents all columns from a table. So it always looks for a table (As we know, there should be FROM clause before specifying table name). So, it throws an error whenever this condition is not satisfied. SELECT COUNT(*) Here, COUNT is a Function. So it is not mandetory to provide a table. Check it out this: DECLARE @cnt INT SET @cnt = COUNT(*) SELECT @cnt SET @cnt = COUNT(‘x’) SELECT @cnt Naveen Select 1 / Select ‘*’ will return 1/* as expected. Select Count(1)/Count(*) will return the count of result set of select statement. Count(1)/Count(*) will have one 1/* for each row in the result set of select statement. Select 1 or Select ‘*’ result set will contain only 1 result. so count is 1. Where as “Select *” is a sysntax which expects the table or equauivalent to table (table functions, etc..). It is like compilation error for that query. Ramesh Hi Friends, Count is an aggregate function and it expects the rows (list of records) for a specified single column or whole rows for *. So, when we use ‘select *’ it definitely give and error because ‘*’ is meant to have all the fields but there is not any table and without table it can only raise an error. So, in the case of ‘Select Count(*)’, there will be an error as a record in the count function so you will get the result as ’1'. Try using : Select COUNT(‘RAMESH’) and think there is an error ‘Must specify table to select from.’ in place of ‘RAMESH’ Pinal : If i am wrong then please clarify this. Sachin Nandanwar Any aggregate function expects a constant or a column name as an expression. DO NOT be confused with * in an aggregate function.The aggregate function does not treat it as a column name or a set of column names but a constant value, as * is a key word in SQL. You can replace any value instead of * for the COUNT function.Ex Select COUNT(5) will result as 1. The error resulting from select * is obvious it expects an object where it can extract the result set. I sincerely thank you all for wonderful conversation, I personally enjoyed it and I am sure all of you have the same feeling. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: CodeProject, Pinal Dave, PostADay, Readers Contribution, Readers Question, SQL, SQL Authority, SQL Puzzle, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, SQLServer, T SQL, Technology

    Read the article

  • SQL SERVER – A Puzzle – Fun with NULL – Fix Error 8117

    - by pinaldave
    During my 8 years of career, I have been involved in many interviews. Quite often, I act as the  interview. If I am the interviewer, I ask many questions – from easy questions to difficult ones. When I am the interviewee, I frequently get an opportunity to ask the interviewer some questions back. Regardless of the my capacity in attending the interview, I always make it a point to ask the interviewer at least one question. What is NULL? It’s always fun to ask this question during interviews, because in every interview, I get a different answer. NULL is often confused with false, absence of value or infinite value. Honestly, NULL is a very interesting subject as it bases its behavior in server settings. There are a few properties of NULL that are universal, but the knowledge about these properties is not known in a universal sense. Let us run this simple puzzle. Run the following T-SQL script: SELECT SUM(data) FROM (SELECT NULL AS data) t It will return the following error: Msg 8117, Level 16, State 1, Line 1 Operand data type NULL is invalid for sum operator. Now the error makes it very clear that NULL is invalid for sum Operator. Frequently enough, I have showed this simple query to many folks whom I came across. I asked them if they could modify the subquery and return the result as NULL. Here is what I expected: Even though this is a very simple looking query, so far I’ve got the correct answer from only 10% of the people to whom I have asked this question. It was common for me to receive this kind of answer – convert the NULL to some data type. However, doing so usually returns the value as 0 or the integer they passed. SELECT SUM(data) FROM (SELECT ISNULL(NULL,0) AS data) t I usually see many people modifying the outer query to get desired NULL result, but that is not allowed in this simple puzzle. This small puzzle made me wonder how many people have a clear understanding about NULL. Well, here is the answer to my simple puzzle. Just CAST NULL AS INT and it will return the final result as NULL: SELECT SUM(data) FROM (SELECT CAST(NULL AS INT) AS data) t Now that you know the answer, don’t you think it was very simple indeed? This blog post is especially dedicated to my friend Madhivanan who has written an excellent blog post about NULL. I am confident that after reading the blog post from Madhivanan, you will have no confusion regarding NULL in the future. Read: NULL, NULL, NULL and nothing but NULL. Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Puzzle, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL SERVER – SSMS: Top Object and Batch Execution Statistics Reports

    - by Pinal Dave
    The month of June till mid of July has been the fever of sports. First, it was Wimbledon Tennis and then the Soccer fever was all over. There is a huge number of fan followers and it is great to see the level at which people sometimes worship these sports. Being an Indian, I cannot forget to mention the India tour of England later part of July. Following these sports and as the events unfold to the finals, there are a number of ways the statisticians can slice and dice the numbers. Cue from soccer I can surely say there is a team performance against another team and then there is individual member fairs against a particular opponent. Such statistics give us a fair idea to how a team in the past or in the recent past has fared against each other, head-to-head stats during World cup and during other neutral venue games. All these statistics are just pointers. In reality, they don’t reflect the calibre of the current team because the individuals who performed in each of these games are totally different (Typical example being the Brazil Vs Germany semi-final match in FIFA 2014). So at times these numbers are misleading. It is worth investigating and get the next level information. Similar to these statistics, SQL Server Management studio is also equipped with a number of reports like a) Object Execution Statistics report and b) Batch Execution Statistics reports. As discussed in the example, the team scorecard is like the Batch Execution statistics and individual stats is like Object Level statistics. The analogy can be taken only this far, trust me there is no correlation between SQL Server functioning and playing sports – It is like I think about diet all the time except while I am eating. Performance – Batch Execution Statistics Let us view the first report which can be invoked from Server Node -> Reports -> Standard Reports -> Performance – Batch Execution Statistics. Most of the values that are displayed in this report come from the DMVs sys.dm_exec_query_stats and sys.dm_exec_sql_text(sql_handle). This report contains 3 distinctive sections as outline below.   Section 1: This is a graphical bar graph representation of Average CPU Time, Average Logical reads and Average Logical Writes for individual batches. The Batch numbers are indicative and the details of individual batch is available in section 3 (detailed below). Section 2: This represents a Pie chart of all the batches by Total CPU Time (%) and Total Logical IO (%) by batches. This graphical representation tells us which batch consumed the highest CPU and IO since the server started, provided plan is available in the cache. Section 3: This is the section where we can find the SQL statements associated with each of the batch Numbers. This also gives us the details of Average CPU / Average Logical Reads and Average Logical Writes in the system for the given batch with object details. Expanding the rows, I will also get the # Executions and # Plans Generated for each of the queries. Performance – Object Execution Statistics The second report worth a look is Object Execution statistics. This is a similar report as the previous but turned on its head by SQL Server Objects. The report has 3 areas to look as above. Section 1 gives the Average CPU, Average IO bar charts for specific objects. The section 2 is a graphical representation of Total CPU by objects and Total Logical IO by objects. The final section details the various objects in detail with the Avg. CPU, IO and other details which are self-explanatory. At a high-level both the reports are based on queries on two DMVs (sys.dm_exec_query_stats and sys.dm_exec_sql_text) and it builds values based on calculations using columns in them: SELECT * FROM    sys.dm_exec_query_stats s1 CROSS APPLY sys.dm_exec_sql_text(sql_handle) AS s2 WHERE   s2.objectid IS NOT NULL AND DB_NAME(s2.dbid) IS NOT NULL ORDER BY  s1.sql_handle; This is one of the simplest form of reports and in future blogs we will look at more complex reports. I truly hope that these reports can give DBAs and developers a hint about what is the possible performance tuning area. As a closing point I must emphasize that all above reports pick up data from the plan cache. If a particular query has consumed a lot of resources earlier, but plan is not available in the cache, none of the above reports would show that bad query. Reference: Pinal Dave (http://blog.sqlauthority.com)Filed under: SQL, SQL Authority, SQL Query, SQL Server, SQL Server Management Studio, SQL Tips and Tricks, T SQL Tagged: SQL Reports

    Read the article

  • SQL SERVER – Updating Data in A Columnstore Index

    - by pinaldave
    So far I have written two articles on Columnstore Indexes, and both of them got very interesting readership. In fact, just recently I got a query on my previous article on Columnstore Index. Read the following two articles to get familiar with the Columnstore Index. They will give you a reference to the question which was asked by a certain reader: SQL SERVER – Fundamentals of Columnstore Index SQL SERVER – How to Ignore Columnstore Index Usage in Query Here is the reader’s question: ” When I tried to update my table after creating the Columnstore index, it gives me an error. What should I do?” When the Columnstore index is created on the table, the table becomes Read-Only table and it does not let any insert/update/delete on the table. The basic understanding is that Columnstore Index will be created on the table that is very huge and holds lots of data. If a table is small enough, there is no need to create a Columnstore index. The regular index should just help it. The reason why Columnstore index was needed is because the table was so big that retrieving the data was taking a really, really long time. Now, updating such a huge table is always a challenge by itself. If the Columnstore Index is created on the table, and the table needs to be updated, you need to know that there are various ways to update it. The easiest way is to disable the Index and enable it. Consider the following code: USE AdventureWorks GO -- Create New Table CREATE TABLE [dbo].[MySalesOrderDetail]( [SalesOrderID] [int] NOT NULL, [SalesOrderDetailID] [int] NOT NULL, [CarrierTrackingNumber] [nvarchar](25) NULL, [OrderQty] [smallint] NOT NULL, [ProductID] [int] NOT NULL, [SpecialOfferID] [int] NOT NULL, [UnitPrice] [money] NOT NULL, [UnitPriceDiscount] [money] NOT NULL, [LineTotal] [numeric](38, 6) NOT NULL, [rowguid] [uniqueidentifier] NOT NULL, [ModifiedDate] [datetime] NOT NULL ) ON [PRIMARY] GO -- Create clustered index CREATE CLUSTERED INDEX [CL_MySalesOrderDetail] ON [dbo].[MySalesOrderDetail] ( [SalesOrderDetailID]) GO -- Create Sample Data Table -- WARNING: This Query may run upto 2-10 minutes based on your systems resources INSERT INTO [dbo].[MySalesOrderDetail] SELECT S1.* FROM Sales.SalesOrderDetail S1 GO 100 -- Create ColumnStore Index CREATE NONCLUSTERED COLUMNSTORE INDEX [IX_MySalesOrderDetail_ColumnStore] ON [MySalesOrderDetail] (UnitPrice, OrderQty, ProductID) GO -- Attempt to Update the table UPDATE [dbo].[MySalesOrderDetail] SET OrderQty = OrderQty +1 WHERE [SalesOrderID] = 43659 GO /* It will throw following error Msg 35330, Level 15, State 1, Line 2 UPDATE statement failed because data cannot be updated in a table with a columnstore index. Consider disabling the columnstore index before issuing the UPDATE statement, then rebuilding the columnstore index after UPDATE is complete. */ A similar error also shows up for Insert/Delete function. Here is the workaround. Disable the Columnstore Index and performance update, enable the Columnstore Index: -- Disable the Columnstore Index ALTER INDEX [IX_MySalesOrderDetail_ColumnStore] ON [dbo].[MySalesOrderDetail] DISABLE GO -- Attempt to Update the table UPDATE [dbo].[MySalesOrderDetail] SET OrderQty = OrderQty +1 WHERE [SalesOrderID] = 43659 GO -- Rebuild the Columnstore Index ALTER INDEX [IX_MySalesOrderDetail_ColumnStore] ON [dbo].[MySalesOrderDetail] REBUILD GO This time it will not throw an error while the update of the table goes successfully. Let us do a cleanup of our tables using this code: -- Cleanup DROP INDEX [IX_MySalesOrderDetail_ColumnStore] ON [dbo].[MySalesOrderDetail] GO TRUNCATE TABLE dbo.MySalesOrderDetail GO DROP TABLE dbo.MySalesOrderDetail GO In the next post we will see how we can use Partition to update the Columnstore Index. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Index, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL SERVER – Read Only Files and SQL Server Management Studio (SSMS)

    - by pinaldave
    Just like any other Developer or DBA SQL Server Management Studio is my favorite application. Any any moment of the time I have multiple instances of the same application are open and I am working on it. Recently, I have come across a very interesting feature in SSMS related to “Read Only” files. I believe it is a little unknown feature as well so decided to write a blog about the same. First create a read only SQL file. You can make any file read by Right Click >> Properties >> Select Attribute Read Only. Now open the same file in SQL Server Management Studio. You will find that besides the file name there is a small ‘lock’ icon. This small icon indicates that the file is read only. Now let us attempt to edit the read only file. It will let us edit the file any way we want, however when we attempt to save it, it gives following pop-up value. The options in the pop-up are self explanatory and I liked it. The goal of the read only file is to prevent users to make un-intended changes. However, when a user should have complete control over the user file. User should be aware that the file is read only but if he wants to edit the file or save as a new file the choices should be present in front of it and the pop-up menu precisely captures the same. Now let us check option related to this feature in SSMS. Go to Menu >> Options >> Environment >> Documents You will find the third option which is “Allow editing of read-only files; warn when attempt to save”. In the above scenario it was already checked. Let us uncheck the same and do the same exercise which we have done earlier. I closed all the earlier window to avoid confusion. With the new option selected when I attempt to even modify the Read Only file, it gives me totally different pop up screen. It gives me an option like “Edit In-Memory”, “Make Writeable” etc. When you select “Edit In-Memory” it allows you to edit the file and later you can save as new file – just like the earlier scenario which we have discussed. . If clicked on the Make Writeable it will remove the restriction of the Read Only and file can be edited as pleased. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Server Management Studio, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL SERVER – BACKUPIO, BACKUPBUFFER – Wait Type – Day 14 of 28

    - by pinaldave
    Backup is the most important task for any database admin. Your data is at risk if you are not performing database backup. Honestly, I have seen many DBAs who know how to take backups but do not know how to restore it. (Sigh!) In this blog post we are going to discuss about one of my real experiences with one of my clients – BACKUPIO. When I started to deal with it, I really had no idea how to fix the issue. However, after fixing it at two places, I think I know why this is happening but at the same time, I am not sure the fix is the best solution. The reality is that the fix is not a solution but a workaround (which is not optimal, but get your things done). From Book On-Line: BACKUPIO Occurs when a backup task is waiting for data, or is waiting for a buffer in which to store data. This type is not typical, except when a task is waiting for a tape mount. BACKUPBUFFER Occurs when a backup task is waiting for data, or is waiting for a buffer in which to store data. This type is not typical, except when a task is waiting for a tape mount. BACKUPIO and BACKUPBUFFER Explanation: This wait stats will occur when you are taking the backup on the tape or any other extremely slow backup system. Reducing BACKUPIO and BACKUPBUFFER wait: In my recent consultancy, backup on tape was very slow probably because the tape system was very old. During the time when I explained this wait type reason in the consultancy, the owners immediately decided to replace the tape drive with an alternate system. They had a small SAN enclosure not being used on side, which they decided to re-purpose. After a week, I had received an email from their DBA, saying that the wait stats have reduced drastically. At another location, my client was using a third party tool (please don’t ask me the name of the tool) to take backup. This tool was compressing the backup along with taking backup. I have had a very good experience with this tool almost all the time except this one sparse experience. When I tried to take backup using the native SQL Server compressed backup, there was a very small value on this wait type and the backup was much faster. However, when I attempted with the third party backup tool, this value was very high again and was taking much more time. The third party tool had many other features but the client was not using these features. We end up using the native SQL Server Compressed backup and it worked very well. If I get to see this higher in my future consultancy, I will try to understand this wait type much more in detail and so probably I would able to come to some solid solution. Read all the post in the Wait Types and Queue series. Note: The information presented here is from my experience and there is no way that I claim it to be accurate. I suggest reading Book OnLine for further clarification. All the discussion of Wait Stats in this blog is generic and varies from system to system. It is recommended that you test this on a development server before implementing it to a production server. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, SQL Wait Stats, SQL Wait Types, T SQL, Technology

    Read the article

  • SQL SERVER – Identifying Column Data Type of uniqueidentifier without Querying System Tables

    - by pinaldave
    I love interesting conversations with related to SQL Server. One of my friends Madhivanan always comes up with an interesting point of conversation. Here is one of the conversation between us. I am very confident this blog post will for sure enable you with some new knowledge. Madhi: How do I know if any table has a uniqueidentifier column used in it? Pinal:  I am sure you know that you can do it through some DMV or catalogue views. Madhi: I know that but how can we do that without using DMV or catalogue views? Pinal: Hm… what can I use? Madhi: You can use table name. Pinal: Easy, just say SELECT YourUniqueIdentCol FROM Table. Madhi: Hold on, the question seems to be not clear to you – you do know the name of the column. The matter of the fact, you do not know if the table has uniqueidentifier column. Only information you have is table name. Pinal: Madhi, this seems like you are changing the question when I am close to answer. Madhi: Well, are you clear now? Let me say it again – How do I know if any table has a uniqueidentifier column and what is its value without using any DMV or System Catalogues? Only information you know is table name and you are allowed to return any kind of error if table does not have uniqueidentifier column. Pinal: Do you know the answer? Madhi: Yes. I just wanted to test your knowledge about SQL. Pinal: I will have to think. Let me accept I do not know it right away. Can you share the answer please? Madhi: I won! Here it goes! Pinal: When I have friends like you – who needs enemies? Madhi: (laughter which did not stop for a minute). CREATE TABLE t ( GuidCol UNIQUEIDENTIFIER DEFAULT newsequentialid() ROWGUIDCOL, data VARCHAR(60) ) INSERT INTO t (data) SELECT 'test' INSERT INTO t (data) SELECT 'test1' SELECT $rowguid FROM t DROP TABLE t This is indeed very interesting to me. Please note that this is not the optimal way and there will be many other ways to retrieve uniqueidentifier name and value. What I learned from this was if I am in a rush to check if the table has uniqueidentifier and I do not know the name of the same, I can use SELECT TOP (1) $rowguid and quickly know the name of the column. I can later use the same columnname in my query. Madhi did teach me this new trick. Did you know this? What are other ways to get the check uniqueidentifier column existence in a database? Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Puzzle, SQL Query, SQL Server, SQL Server Management Studio, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL SERVER – Simple Example of Snapshot Isolation – Reduce the Blocking Transactions

    - by pinaldave
    To learn any technology and move to a more advanced level, it is very important to understand the fundamentals of the subject first. Today, we will be talking about something which has been quite introduced a long time ago but not properly explored when it comes to the isolation level. Snapshot Isolation was introduced in SQL Server in 2005. However, the reality is that there are still many software shops which are using the SQL Server 2000, and therefore cannot be able to maintain the Snapshot Isolation. Many software shops have upgraded to the later version of the SQL Server, but their respective developers have not spend enough time to upgrade themselves with the latest technology. “It works!” is a very common answer of many when they are asked about utilizing the new technology, instead of backward compatibility commands. In one of the recent consultation project, I had same experience when developers have “heard about it” but have no idea about snapshot isolation. They were thinking it is the same as Snapshot Replication – which is plain wrong. This is the same demo I am including here which I have created for them. In Snapshot Isolation, the updated row versions for each transaction are maintained in TempDB. Once a transaction has begun, it ignores all the newer rows inserted or updated in the table. Let us examine this example which shows the simple demonstration. This transaction works on optimistic concurrency model. Since reading a certain transaction does not block writing transaction, it also does not block the reading transaction, which reduced the blocking. First, enable database to work with Snapshot Isolation. Additionally, check the existing values in the table from HumanResources.Shift. ALTER DATABASE AdventureWorks SET ALLOW_SNAPSHOT_ISOLATION ON GO SELECT ModifiedDate FROM HumanResources.Shift GO Now, we will need two different sessions to prove this example. First Session: Set Transaction level isolation to snapshot and begin the transaction. Update the column “ModifiedDate” to today’s date. -- Session 1 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN UPDATE HumanResources.Shift SET ModifiedDate = GETDATE() GO Please note that we have not yet been committed to the transaction. Now, open the second session and run the following “SELECT” statement. Then, check the values of the table. Please pay attention on setting the Isolation level for the second one as “Snapshot” at the same time when we already start the transaction using BEGIN TRAN. -- Session 2 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN SELECT ModifiedDate FROM HumanResources.Shift GO You will notice that the values in the table are still original values. They have not been modified yet. Once again, go back to session 1 and begin the transaction. -- Session 1 COMMIT After that, go back to Session 2 and see the values of the table. -- Session 2 SELECT ModifiedDate FROM HumanResources.Shift GO You will notice that the values are yet not changed and they are still the same old values which were there right in the beginning of the session. Now, let us commit the transaction in the session 2. Once committed, run the same SELECT statement once more and see what the result is. -- Session 2 COMMIT SELECT ModifiedDate FROM HumanResources.Shift GO You will notice that it now reflects the new updated value. I hope that this example is clear enough as it would give you good idea how the Snapshot Isolation level works. There is much more to write about an extra level, READ_COMMITTED_SNAPSHOT, which we will be discussing in another post soon. If you wish to use this transaction’s Isolation level in your production database, I would appreciate your comments about their performance on your servers. I have included here the complete script used in this example for your quick reference. ALTER DATABASE AdventureWorks SET ALLOW_SNAPSHOT_ISOLATION ON GO SELECT ModifiedDate FROM HumanResources.Shift GO -- Session 1 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN UPDATE HumanResources.Shift SET ModifiedDate = GETDATE() GO -- Session 2 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN SELECT ModifiedDate FROM HumanResources.Shift GO -- Session 1 COMMIT -- Session 2 SELECT ModifiedDate FROM HumanResources.Shift GO -- Session 2 COMMIT SELECT ModifiedDate FROM HumanResources.Shift GO Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Performance, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology Tagged: Transaction Isolation

    Read the article

  • SQL SERVER – Simple Example of Snapshot Isolation – Reduce the Blocking Transactions

    - by pinaldave
    To learn any technology and move to a more advanced level, it is very important to understand the fundamentals of the subject first. Today, we will be talking about something which has been quite introduced a long time ago but not properly explored when it comes to the isolation level. Snapshot Isolation was introduced in SQL Server in 2005. However, the reality is that there are still many software shops which are using the SQL Server 2000, and therefore cannot be able to maintain the Snapshot Isolation. Many software shops have upgraded to the later version of the SQL Server, but their respective developers have not spend enough time to upgrade themselves with the latest technology. “It works!” is a very common answer of many when they are asked about utilizing the new technology, instead of backward compatibility commands. In one of the recent consultation project, I had same experience when developers have “heard about it” but have no idea about snapshot isolation. They were thinking it is the same as Snapshot Replication – which is plain wrong. This is the same demo I am including here which I have created for them. In Snapshot Isolation, the updated row versions for each transaction are maintained in TempDB. Once a transaction has begun, it ignores all the newer rows inserted or updated in the table. Let us examine this example which shows the simple demonstration. This transaction works on optimistic concurrency model. Since reading a certain transaction does not block writing transaction, it also does not block the reading transaction, which reduced the blocking. First, enable database to work with Snapshot Isolation. Additionally, check the existing values in the table from HumanResources.Shift. ALTER DATABASE AdventureWorks SET ALLOW_SNAPSHOT_ISOLATION ON GO SELECT ModifiedDate FROM HumanResources.Shift GO Now, we will need two different sessions to prove this example. First Session: Set Transaction level isolation to snapshot and begin the transaction. Update the column “ModifiedDate” to today’s date. -- Session 1 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN UPDATE HumanResources.Shift SET ModifiedDate = GETDATE() GO Please note that we have not yet been committed to the transaction. Now, open the second session and run the following “SELECT” statement. Then, check the values of the table. Please pay attention on setting the Isolation level for the second one as “Snapshot” at the same time when we already start the transaction using BEGIN TRAN. -- Session 2 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN SELECT ModifiedDate FROM HumanResources.Shift GO You will notice that the values in the table are still original values. They have not been modified yet. Once again, go back to session 1 and begin the transaction. -- Session 1 COMMIT After that, go back to Session 2 and see the values of the table. -- Session 2 SELECT ModifiedDate FROM HumanResources.Shift GO You will notice that the values are yet not changed and they are still the same old values which were there right in the beginning of the session. Now, let us commit the transaction in the session 2. Once committed, run the same SELECT statement once more and see what the result is. -- Session 2 COMMIT SELECT ModifiedDate FROM HumanResources.Shift GO You will notice that it now reflects the new updated value. I hope that this example is clear enough as it would give you good idea how the Snapshot Isolation level works. There is much more to write about an extra level, READ_COMMITTED_SNAPSHOT, which we will be discussing in another post soon. If you wish to use this transaction’s Isolation level in your production database, I would appreciate your comments about their performance on your servers. I have included here the complete script used in this example for your quick reference. ALTER DATABASE AdventureWorks SET ALLOW_SNAPSHOT_ISOLATION ON GO SELECT ModifiedDate FROM HumanResources.Shift GO -- Session 1 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN UPDATE HumanResources.Shift SET ModifiedDate = GETDATE() GO -- Session 2 SET TRANSACTION ISOLATION LEVEL SNAPSHOT BEGIN TRAN SELECT ModifiedDate FROM HumanResources.Shift GO -- Session 1 COMMIT -- Session 2 SELECT ModifiedDate FROM HumanResources.Shift GO -- Session 2 COMMIT SELECT ModifiedDate FROM HumanResources.Shift GO Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Performance, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology Tagged: Transaction Isolation

    Read the article

  • SQL SERVER – LCK_M_XXX – Wait Type – Day 15 of 28

    - by pinaldave
    Locking is a mechanism used by the SQL Server Database Engine to synchronize access by multiple users to the same piece of data, at the same time. In simpler words, it maintains the integrity of data by protecting (or preventing) access to the database object. From Book On-Line: LCK_M_BU Occurs when a task is waiting to acquire a Bulk Update (BU) lock. LCK_M_IS Occurs when a task is waiting to acquire an Intent Shared (IS) lock. LCK_M_IU Occurs when a task is waiting to acquire an Intent Update (IU) lock. LCK_M_IX Occurs when a task is waiting to acquire an Intent Exclusive (IX) lock. LCK_M_S Occurs when a task is waiting to acquire a Shared lock. LCK_M_SCH_M Occurs when a task is waiting to acquire a Schema Modify lock. LCK_M_SCH_S Occurs when a task is waiting to acquire a Schema Share lock. LCK_M_SIU Occurs when a task is waiting to acquire a Shared With Intent Update lock. LCK_M_SIX Occurs when a task is waiting to acquire a Shared With Intent Exclusive lock. LCK_M_U Occurs when a task is waiting to acquire an Update lock. LCK_M_UIX Occurs when a task is waiting to acquire an Update With Intent Exclusive lock. LCK_M_X Occurs when a task is waiting to acquire an Exclusive lock. LCK_M_XXX Explanation: I think the explanation of this wait type is the simplest. When any task is waiting to acquire lock on any resource, this particular wait type occurs. The common reason for the task to be waiting to put lock on the resource is that the resource is already locked and some other operations may be going on within it. This wait also indicates that resources are not available or are occupied at the moment due to some reasons. There is a good chance that the waiting queries start to time out if this wait type is very high. Client application may degrade the performance as well. You can use various methods to find blocking queries: EXEC sp_who2 SQL SERVER – Quickest Way to Identify Blocking Query and Resolution – Dirty Solution DMV – sys.dm_tran_locks DMV – sys.dm_os_waiting_tasks Reducing LCK_M_XXX wait: Check the Explicit Transactions. If transactions are very long, this wait type can start building up because of other waiting transactions. Keep the transactions small. Serialization Isolation can build up this wait type. If that is an acceptable isolation for your business, this wait type may be natural. The default isolation of SQL Server is ‘Read Committed’. One of my clients has changed their isolation to “Read Uncommitted”. I strongly discourage the use of this because this will probably lead to having lots of dirty data in the database. Identify blocking queries mentioned using various methods described above, and then optimize them. Partition can be one of the options to consider because this will allow transactions to execute concurrently on different partitions. If there are runaway queries, use timeout. (Please discuss this solution with your database architect first as timeout can work against you). Check if there is no memory and IO-related issue using the following counters: Checking Memory Related Perfmon Counters SQLServer: Memory Manager\Memory Grants Pending (Consistent higher value than 0-2) SQLServer: Memory Manager\Memory Grants Outstanding (Consistent higher value, Benchmark) SQLServer: Buffer Manager\Buffer Hit Cache Ratio (Higher is better, greater than 90% for usually smooth running system) SQLServer: Buffer Manager\Page Life Expectancy (Consistent lower value than 300 seconds) Memory: Available Mbytes (Information only) Memory: Page Faults/sec (Benchmark only) Memory: Pages/sec (Benchmark only) Checking Disk Related Perfmon Counters Average Disk sec/Read (Consistent higher value than 4-8 millisecond is not good) Average Disk sec/Write (Consistent higher value than 4-8 millisecond is not good) Average Disk Read/Write Queue Length (Consistent higher value than benchmark is not good) Read all the post in the Wait Types and Queue series. Note: The information presented here is from my experience and there is no way that I claim it to be accurate. I suggest reading Book OnLine for further clarification. All the discussion of Wait Stats in this blog is generic and varies from system to system. It is recommended that you test this on a development server before implementing it to a production server. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, SQL Wait Stats, SQL Wait Types, T SQL, Technology

    Read the article

  • SQL SERVER – Size of Index Table for Each Index – Solution 3 – Powershell

    - by pinaldave
    Laerte Junior If you are a Powershell user, the name of the Laerte Junior is not a new name. He is the one man with exceptional knowledge of Powershell. He is not only very knowledgeable, but also very kind and eager to those in need. I have been attempting to setup Powershell for many days, but constantly facing issues. I was not able to get going with this tool. Finally, yesterday I sent email to Laerte in response to his comment posted here. Within 5 minutes, Laerte came online and helped me with the solution. He spend nearly 15 minutes working along with me to solve my problem with installation. And yes, he did resolve it remotely without looking at my screen – What a skilled and exceptional person!! I will soon post a detail note about the issue I faced and resolved with the help of Laerte. Here is his solution to my earlier puzzle in his own words. Read the original puzzle here and Laerte’s solution from here. Hi Pinal, I do not say better, but maybe another approach to enthusiasts in powershell and SQLSPX library would be: 1 – All indexes in all tables and all databases Get-SqlDatabase -sqlserver “Yourserver” | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused 2 – All Indexes in all tables and specific database Get-SqlDatabase -sqlserver “Yourserver” “Yourdb” | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused 3 – All Indexes in specific table and database Get-SqlDatabase -sqlserver “Yourserver” “Yourdb” | Get-SqlTable “YourTable” | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused and to output to txt.. pipe Out-File Get-SqlDatabase -sqlserver “Yourserver” | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused | out-file c:\IndexesSize.txt If you have one txt with all your servers, can be for all of them also. Lets say you have all your servers in servers.txt: something like NameServer1 NameServer2 NameServer3 NameServer4 We could Use : foreach ($Server in Get-content c:\temp\servers.txt) { Get-SqlDatabase -sqlserver $Server | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused } :) After fixing my issue with Powershell, I ran Laerte‘s second suggestion – “All Indexes in all tables and specific database” and found the following accurate output. Click to Enlarge Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Index, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology Tagged: Powershell

    Read the article

  • SQL SERVER – Size of Index Table for Each Index – Solution 3 – Powershell

    - by pinaldave
    Laerte Junior If you are a Powershell user, the name of the Laerte Junior is not a new name. He is the one man with exceptional knowledge of Powershell. He is not only very knowledgeable, but also very kind and eager to those in need. I have been attempting to setup Powershell for many days, but constantly facing issues. I was not able to get going with this tool. Finally, yesterday I sent email to Laerte in response to his comment posted here. Within 5 minutes, Laerte came online and helped me with the solution. He spend nearly 15 minutes working along with me to solve my problem with installation. And yes, he did resolve it remotely without looking at my screen – What a skilled and exceptional person!! I will soon post a detail note about the issue I faced and resolved with the help of Laerte. Here is his solution to my earlier puzzle in his own words. Read the original puzzle here and Laerte’s solution from here. Hi Pinal, I do not say better, but maybe another approach to enthusiasts in powershell and SQLSPX library would be: 1 – All indexes in all tables and all databases Get-SqlDatabase -sqlserver “Yourserver” | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused 2 – All Indexes in all tables and specific database Get-SqlDatabase -sqlserver “Yourserver” “Yourdb” | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused 3 – All Indexes in specific table and database Get-SqlDatabase -sqlserver “Yourserver” “Yourdb” | Get-SqlTable “YourTable” | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused and to output to txt.. pipe Out-File Get-SqlDatabase -sqlserver “Yourserver” | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused | out-file c:\IndexesSize.txt If you have one txt with all your servers, can be for all of them also. Lets say you have all your servers in servers.txt: something like NameServer1 NameServer2 NameServer3 NameServer4 We could Use : foreach ($Server in Get-content c:\temp\servers.txt) { Get-SqlDatabase -sqlserver $Server | Get-SqlTable | Get-SqlIndex | Format-table Server,dbname,schema,table,name,id,spaceused } :) After fixing my issue with Powershell, I ran Laerte‘s second suggestion – “All Indexes in all tables and specific database” and found the following accurate output. Click to Enlarge Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Index, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology Tagged: Powershell

    Read the article

  • SQL SERVER – SSMS Automatically Generates TOP (100) PERCENT in Query Designer

    - by pinaldave
    Earlier this week, I was surfing various SQL forums to see what kind of help developer need in the SQL Server world. One of the question indeed caught my attention. I am here regenerating complete question as well scenario to illustrate the point in a precise manner. Additionally, I have added added second part of the question to give completeness. Question: I am trying to create a view in Query Designer (not in the New Query Window). Every time I am trying to create a view it always adds  TOP (100) PERCENT automatically on the T-SQL script. No matter what I do, it always automatically adds the TOP (100) PERCENT to the script. I have attempted to copy paste from notepad, build a query and a few other things – there is no success. I am really not sure what I am doing wrong with Query Designer. Here is my query script: (I use AdventureWorks as a sample database) SELECT Person.Address.AddressID FROM Person.Address INNER JOIN Person.AddressType ON Person.Address.AddressID = Person.AddressType.AddressTypeID ORDER BY Person.Address.AddressID This script automatically replaces by following query: SELECT TOP (100) PERCENT Person.Address.AddressID FROM Person.Address INNER JOIN Person.AddressType ON Person.Address.AddressID = Person.AddressType.AddressTypeID ORDER BY Person.Address.AddressID However, when I try to do the same from New Query Window it works totally fine. However, when I attempt to create a view of the same query it gives following error. Msg 1033, Level 15, State 1, Procedure myView, Line 6 The ORDER BY clause is invalid in views, inline functions, derived tables, subqueries, and common table expressions, unless TOP, OFFSET or FOR XML is also specified. It is pretty clear to me now that the script which I have written seems to need TOP (100) PERCENT, so Query . Why do I need it? Is there any work around to this issue. I particularly find this question pretty interesting as it really touches the fundamentals of the T-SQL query writing. Please note that the query which is automatically changed is not in New Query Editor but opened from SSMS using following way. Database >> Views >> Right Click >> New View (see the image below) Answer: The answer to the above question can be very long but I will keep it simple and to the point. There are three things to discuss in above script 1) Reason for Error 2) Reason for Auto generates TOP (100) PERCENT and 3) Potential solutions to the above error. Let us quickly see them in detail. 1) Reason for Error The reason for error is already given in the error. ORDER BY is invalid in the views and a few other objects. One has to use TOP or other keywords along with it. The way semantics of the query works where optimizer only follows(honors) the ORDER BY in the same scope or the same SELECT/UPDATE/DELETE statement. There is a possibility that one can order after the scope of the view again the efforts spend to order view will be wasted. The final resultset of the query always follows the final ORDER BY or outer query’s order and due to the same reason optimizer follows the final order of the query and not of the views (as view will be used in another query for further processing e.g. in SELECT statement). Due to same reason ORDER BY is now allowed in the view. For further accuracy and clear guidance I suggest you read this blog post by Query Optimizer Team. They have explained it very clear manner the same subject. 2) Reason for Auto Generated TOP (100) PERCENT One of the most popular workaround to above error is to use TOP (100) PERCENT in the view. Now TOP (100) PERCENT allows user to use ORDER BY in the query and allows user to overcome above error which we discussed. This gives the impression to the user that they have resolved the error and successfully able to use ORDER BY in the View. Well, this is incorrect as well. The way this works is when TOP (100) PERCENT is used the result is not guaranteed as well it is ignored in our the query where the view is used. Here is the blog post on this subject: Interesting Observation – TOP 100 PERCENT and ORDER BY. Now when you create a new view in the SSMS and build a query with ORDER BY to avoid the error automatically it adds the TOP 100 PERCENT. Here is the connect item for the same issue. I am sure there will be more connect items as well but I could not find them. 3) Potential Solutions If you are reading this post from the beginning in that case, it is clear by now that ORDER BY should not be used in the View as it does not serve any purpose unless there is a specific need of it. If you are going to use TOP 100 PERCENT with ORDER BY there is absolutely no need of using ORDER BY rather avoid using it all together. Here is another blog post of mine which describes the same subject ORDER BY Does Not Work – Limitation of the Views Part 1. It is valid to use ORDER BY in a view if there is a clear business need of using TOP with any other percentage lower than 100 (for example TOP 10 PERCENT or TOP 50 PERCENT etc). In most of the cases ORDER BY is not needed in the view and it should be used in the most outer query for present result in desired order. User can remove TOP 100 PERCENT and ORDER BY from the view before using the view in any query or procedure. In the most outer query there should be ORDER BY as per the business need. I think this sums up the concept in a few words. This is a very long topic and not easy to illustrate in one single blog post. I welcome your comments and suggestions. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Server Management Studio, SQL Tips and Tricks, SQL View, T SQL, Technology

    Read the article

  • SQLAuthority News – Download Pre-configured VHD – SQL Server 2008 R2 Standard on Windows Server 2008 R2 SP1 Standard

    - by pinaldave
    It is extremely simple to test out latest SQL Server 2008 R2. You can even get pre-configured ready to use VHD, which you can download and use it. This becomes very easy as one does not have to do anything besides downloading VHD and installing it on Hyper-V. Download Pre-configured VHD – SQL Server 2008 R2 Standard provides a trusted, productive and intelligent data platform that enables you to run your most demanding mission-critical applications, reduce time and cost of development and management of applications, and deliver actionable insight to your entire organization. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Documentation, SQL Download, SQL Query, SQL Server, SQL Tips and Tricks, SQLAuthority News, T SQL, Technology

    Read the article

  • SQL SERVER – Puzzle Involving NULL – Resolve – Error – Operand data type void type is invalid for sum operator

    - by pinaldave
    Today is Monday let us start this week with interesting puzzle. Yesterday I had also posted quick question here: SQL SERVER – T-SQL Scripts to Find Maximum between Two Numbers Run following code: SELECT SUM(data) FROM (SELECT NULL AS DATA) t It will throw following error. Msg 8117, Level 16, State 1, Line 1 Operand data type void type is invalid for sum operator. I can easily fix if I use ISNULL Function as displayed following. SELECT SUM(data) FROM (SELECT ISNULL(NULL,0) AS DATA) t Above script will not throw an error. However, there is one more method how this can be fixed. Can you come up with another method which will not generate error? Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Puzzle, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, SQLServer, T SQL, Technology

    Read the article

  • SQL SERVER – SSMS: Database Consistency History Report

    - by Pinal Dave
    Doctor and Database The last place I like to visit is always a hospital. With the monsoon season starting, intermittent rains, it has become sort of a routine to get a cycle of fever every other year (seriously I hate it). So when I visit my doctor, it is always interesting in the way he quizzes me. The routine question of – “How many days have you had this?”, “Is there any pattern?”, “Did you drench in rain?”, “Do you have any other symptom?” and so on. The idea here is that the doctor wants to find any anomaly or a pattern that will guide him to a viral or bacterial type. Most of the time they get it based on experience and sometimes after a battery of tests. So if there is consistent behavior to your problem, there is always a solution out. SQL Server has its way to find if the server data / files are in consistent state using the DBCC commands. Back to SQL Server In real life, Database consistency check is one of the critical operations a DBA generally doesn’t give much priority. Many readers of my blogs have asked many times, how do we know if the database is consistent? How do I read output of DBCC CHECKDB and find if everything is right or not? My common answer to all of them is – look at the bottom of checkdb (or checktable) output and look for below line. CHECKDB found 0 allocation errors and 0 consistency errors in database ‘DatabaseName’. Above is a “good sign” because we are seeing zero allocation and zero consistency error. If you are seeing non-zero errors then there is some problem with the database. Sample output is shown as below: CHECKDB found 0 allocation errors and 2 consistency errors in database ‘DatabaseName’. repair_allow_data_loss is the minimum repair level for the errors found by DBCC CHECKDB (DatabaseName). If we see non-zero error then most of the time (not always) we get repair options depending on the level of corruption. There is risk involved with above option (repair_allow_data_loss), that is – we would lose the data. Sometimes the option would be repair_rebuild which is little safer. Though these options are available, it is important to find the root cause to the problem. In standard report, there is a report which can show the history of checkdb executed for the selected database. Since this is a database level report, we need to right click on database, click Reports, click Standard Reports and then choose “Database Consistency History” report. The information in this report is picked from default trace. If default trace is disabled or there is no checkdb run or information is not there in default trace (because it’s rolled over), we would get report like below. As we can see report says it very clearly: Currently, no execution history of CHECKDB is available or default trace is not enabled. To demonstrate, I have caused corruption in one of the database and did below steps. Run CheckDB so that errors are reported. Fix the corruption by losing the data using repair option Run CheckDB again to check if corruption is cleared. After that I have launched the report and below is what we would see. If you are lazy like me and don’t want to run the report manually for each database then below query would be handy to provide same report for all database. This query is runs behind the scenes by the report. All I have done is remove the filter for database name (at the last – highlighted). DECLARE @curr_tracefilename VARCHAR(500); DECLARE @base_tracefilename VARCHAR(500); DECLARE @indx INT; SELECT @curr_tracefilename = path FROM sys.traces WHERE is_default = 1; SET @curr_tracefilename = REVERSE(@curr_tracefilename); SELECT @indx  = PATINDEX('%\%', @curr_tracefilename) ; SET @curr_tracefilename = REVERSE(@curr_tracefilename); SET @base_tracefilename = LEFT( @curr_tracefilename,LEN(@curr_tracefilename) - @indx) + '\log.trc'; SELECT  SUBSTRING(CONVERT(NVARCHAR(MAX),TEXTData),36, PATINDEX('%executed%',TEXTData)-36) AS command ,       LoginName ,       StartTime ,       CONVERT(INT,SUBSTRING(CONVERT(NVARCHAR(MAX),TEXTData),PATINDEX('%found%',TEXTData) +6,PATINDEX('%errors %',TEXTData)-PATINDEX('%found%',TEXTData)-6)) AS errors ,       CONVERT(INT,SUBSTRING(CONVERT(NVARCHAR(MAX),TEXTData),PATINDEX('%repaired%',TEXTData) +9,PATINDEX('%errors.%',TEXTData)-PATINDEX('%repaired%',TEXTData)-9)) repaired ,       SUBSTRING(CONVERT(NVARCHAR(MAX),TEXTData),PATINDEX('%time:%',TEXTData)+6,PATINDEX('%hours%',TEXTData)-PATINDEX('%time:%',TEXTData)-6)+':'+SUBSTRING(CONVERT(NVARCHAR(MAX),TEXTData),PATINDEX('%hours%',TEXTData) +6,PATINDEX('%minutes%',TEXTData)-PATINDEX('%hours%',TEXTData)-6)+':'+SUBSTRING(CONVERT(NVARCHAR(MAX),TEXTData),PATINDEX('%minutes%',TEXTData) +8,PATINDEX('%seconds.%',TEXTData)-PATINDEX('%minutes%',TEXTData)-8) AS time FROM::fn_trace_gettable( @base_tracefilename, DEFAULT) WHERE EventClass = 22 AND SUBSTRING(TEXTData,36,12) = 'DBCC CHECKDB' -- AND DatabaseName = @DatabaseName; Don’t get worried about the logic above. All it is doing is reading the trace files, parsing below entry and getting out information for underlined words. DBCC CHECKDB (CorruptedDatabase) executed by sa found 2 errors and repaired 0 errors. Elapsed time: 0 hours 0 minutes 0 seconds.  Internal database snapshot has split point LSN = 00000029:00000030:0001 and first LSN = 00000029:00000020:0001. Hopefully now onwards you would run checkdb and understand the importance of it. As responsible DBAs I am sure you are already doing it, let me know how often do you actually run them on you production environment? Reference: Pinal Dave (http://blog.sqlauthority.com)Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Server Management Studio, SQL Tips and Tricks, T SQL Tagged: SQL Reports

    Read the article

  • SQL SERVER – SSMS: Disk Usage Report

    - by Pinal Dave
    Let us start with humor!  I think we the series on various reports, we come to a logical point. We covered all the reports at server level. This means the reports we saw were targeted towards activities that are related to instance level operations. These are mostly like how a doctor diagnoses a patient. At this point I am reminded of a dialog which I read somewhere: Patient: Doc, It hurts when I touch my head. Doc: Ok, go on. What else have you experienced? Patient: It hurts even when I touch my eye, it hurts when I touch my arms, it even hurts when I touch my feet, etc. Doc: Hmmm … Patient: I feel it hurts when I touch anywhere in my body. Doc: Ahh … now I get it. You need a plaster to your finger John. Sometimes the server level gives an indicator to what is happening in the system, but we need to get to the root cause for a specific database. So, this is the first blog in series where we would start discussing about database level reports. To launch database level reports, expand selected server in Object Explorer, expand the Databases folder, and then right-click any database for which we want to look at reports. From the menu, select Reports, then Standard Reports, and then any of database level reports. In this blog, we would talk about four “disk” reports because they are similar: Disk Usage Disk Usage by Top Tables Disk Usage by Table Disk Usage by Partition Disk Usage This report shows multiple information about the database. Let us discuss them one by one.  We have divided the output into 5 different sections. Section 1 shows the high level summary of the database. It shows the space used by database files (mdf and ldf). Under the hood, the report uses, various DMVs and DBCC Commands, it is using sys.data_spaces and DBCC SHOWFILESTATS. Section 2 and 3 are pie charts. One for data file allocation and another for the transaction log file. Pie chart for “Data Files Space Usage (%)” shows space consumed data, indexes, allocated to the SQL Server database, and unallocated space which is allocated to the SQL Server database but not yet filled with anything. “Transaction Log Space Usage (%)” used DBCC SQLPERF (LOGSPACE) and shows how much empty space we have in the physical transaction log file. Section 4 shows the data from Default Trace and looks at Event IDs 92, 93, 94, 95 which are for “Data File Auto Grow”, “Log File Auto Grow”, “Data File Auto Shrink” and “Log File Auto Shrink” respectively. Here is an expanded view for that section. If default trace is not enabled, then this section would be replaced by the message “Trace Log is disabled” as highlighted below. Section 5 of the report uses DBCC SHOWFILESTATS to get information. Here is the enhanced version of that section. This shows the physical layout of the file. In case you have In-Memory Objects in the database (from SQL Server 2014), then report would show information about those as well. Here is the screenshot taken for a different database, which has In-Memory table. I have highlighted new things which are only shown for in-memory database. The new sections which are highlighted above are using sys.dm_db_xtp_checkpoint_files, sys.database_files and sys.data_spaces. The new type for in-memory OLTP is ‘FX’ in sys.data_space. The next set of reports is targeted to get information about a table and its storage. These reports can answer questions like: Which is the biggest table in the database? How many rows we have in table? Is there any table which has a lot of reserved space but its unused? Which partition of the table is having more data? Disk Usage by Top Tables This report provides detailed data on the utilization of disk space by top 1000 tables within the Database. The report does not provide data for memory optimized tables. Disk Usage by Table This report is same as earlier report with few difference. First Report shows only 1000 rows First Report does order by values in DMV sys.dm_db_partition_stats whereas second one does it based on name of the table. Both of the reports have interactive sort facility. We can click on any column header and change the sorting order of data. Disk Usage by Partition This report shows the distribution of the data in table based on partition in the table. This is so similar to previous output with the partition details now. Here is the query taken from profiler. SELECT row_number() OVER (ORDER BY a1.used_page_count DESC, a1.index_id) AS row_number ,      (dense_rank() OVER (ORDER BY a5.name, a2.name))%2 AS l1 ,      a1.OBJECT_ID ,      a5.name AS [schema] ,       a2.name ,       a1.index_id ,       a3.name AS index_name ,       a3.type_desc ,       a1.partition_number ,       a1.used_page_count * 8 AS total_used_pages ,       a1.reserved_page_count * 8 AS total_reserved_pages ,       a1.row_count FROM sys.dm_db_partition_stats a1 INNER JOIN sys.all_objects a2  ON ( a1.OBJECT_ID = a2.OBJECT_ID) AND a1.OBJECT_ID NOT IN (SELECT OBJECT_ID FROM sys.tables WHERE is_memory_optimized = 1) INNER JOIN sys.schemas a5 ON (a5.schema_id = a2.schema_id) LEFT OUTER JOIN  sys.indexes a3  ON ( (a1.OBJECT_ID = a3.OBJECT_ID) AND (a1.index_id = a3.index_id) ) WHERE (SELECT MAX(DISTINCT partition_number) FROM sys.dm_db_partition_stats a4 WHERE (a4.OBJECT_ID = a1.OBJECT_ID)) >= 1 AND a2.TYPE <> N'S' AND  a2.TYPE <> N'IT' ORDER BY a5.name ASC, a2.name ASC, a1.index_id, a1.used_page_count DESC, a1.partition_number Using all of the above reports, you should be able to get the usage of database files and also space used by tables. I think this is too much disk information for a single blog and I hope you have used them in the past to get data. Do let me know if you found anything interesting using these reports in your environments. Reference: Pinal Dave (http://blog.sqlauthority.com)Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Server Management Studio, SQL Tips and Tricks, T SQL Tagged: SQL Reports

    Read the article

  • SQL SERVER – Solution – 2 T-SQL Puzzles – Display Star and Shortest Code to Display 1

    - by pinaldave
    Earlier on this blog we had asked two puzzles. The response from all of you is nothing but Amazing. I have received 350+ responses. Many are valid and many were indeed something I had not thought about it. I strongly suggest you read all the puzzles and their answers here - trust me if you start reading the comments you will not stop till you read every single comment. Seriously trust me on it. Personally I have learned a lot from it. Let us recap the puzzles here quickly. Puzzle 1: Why following code when executed in SSMS displays result as a * (Star)? SELECT CAST(634 AS VARCHAR(2)) Puzzle 2: Write the shortest code that produces results as 1 without using any numbers in the select statement. Bonus Q: How many different Operating System (OS) NuoDB support? As I mentioned earlier the participation was nothing but Amazing. I will write about the winners and the best answers in short time. Meanwhile I will give to the point answers to above puzzles. Solution 1: When you convert character or binary expressions (char, nchar, nvarchar, varchar,binary, or varbinary) to an expression of a different data type, data can be truncated, only partially displayed, or an error is returned because the result is too short to display. Conversions to char, varchar, nchar, nvarchar, binary, and varbinary are truncated, except for the conversions shown in the following table. Reference of the text and table from MSDN. Solution 2: The shortest code to produce answer 1 : SELECT EXP($) or SELECT COS($) or SELECT DAY($) When SELECT $ it gives us the result as 0.00 and the EXP of the same is 1. I believe it is pretty neat. There were plenty other answers but this was the shortest. Another shorter answer would be PRINT EXP($) but no one has proposed that as in original Question I have explicitly mentioned SELECT in the original question. Bonus Answer: 5 OS: Windows, MacOS, Linux, Solaris, Joyent SmartOS Reference Please do read every single comment here. Do leave a comment which one do you think is the best comment out of all the comments. Meanwhile if there is a better solution and I have missed it do let me know as we still have time to correct it. I will be selecting the winner before the weekend as I am going through each and every of 350 comment. I will be selecting the best comments along with the winning comment. If our selection matches – one of you may still win something cool.  Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Puzzle, SQL Query, SQL Server, SQL Tips and Tricks, SQLServer, T SQL, Technology Tagged: NuoDB

    Read the article

  • SQL SERVER – master Database Log File Grew Too Big

    - by pinaldave
    Couple of the days ago, I received following email and I find this email very interesting and I feel like sharing with all of you. Note: Please read the whole email before providing your suggestions. “Hi Pinal, If you can share these details on your blog, it will help many. We understand the value of the master database and we take its regular back up (everyday midnight). Yesterday we noticed that our master database log file has grown very large. This is very first time that we have encountered such an issue. The master database is in simple recovery mode; so we assumed that it will never grow big; however, we now have a big log file. We ran the following command USE [master] GO DBCC SHRINKFILE (N'mastlog' , 0, TRUNCATEONLY) GO We know this command will break the chains of LSN but as per our understanding; it should not matter as we are in simple recovery model.     After running this, the log file becomes very small. Just to be cautious, we took full backup of the master database right away. We totally understand that this is not the normal practice; so if you are going to tell us the same, we are aware of it. However, here is the question for you? What operation in master database would have caused our log file to grow too large? Thanks, [name and company name removed as per request]“ Here was my response to them: “Hi [name removed], It is great that you are aware of all the right steps and method. Taking full backup when you are not sure is always a good practice. Regarding your question what could have caused your master database log to grow larger, let me try to guess what could have happened. Do you have any user table in the master database? If yes, this is not recommended and also NOT a good practice. If have user tables in master database and you are doing any long operation (may be lots of insert, update, delete or rebuilding them), then it can cause this situation. You have made me curious about your scenario; do revert back. Kind Regards, Pinal” Within few minutes I received reply: “That was it Pinal. We had one of the maintenance task log tables created in the master table, which had many long transactions during the night. We moved it to newly created database named ‘maintenance’, and we will keep you updated.” I was very glad to receive the email. I do not suggest that any user table should be created in the master database. It should be left alone from user objects. Now here is the question for you – can you think of any other reason for master log file growth? Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Backup and Restore, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQLAuthority News – SQL Server 2008 R2 Hosted Trial

    - by pinaldave
    This is a bit old news but for me but it will new for many of you know. SQLPASS, Dell, Microsoft and MaximumASP has come together and build hosted environment for free to all of us to use and experiment with. Register now to try out up to seven labs: SQL Server 2008 R2 – Multi Server Management SQL Server 2008 R2 – PowerPivot SQL Server 2008 R2 – Reporting Services SQL Server 2008 R2 – Master Data Services SQL Server 2008 R2 – StreamInsight SQL Server Integration Services – Introduction SQL Server Integration Services – Intermediate to Advanced Now this is indeed wonderful opportunity as you do not need to buy anything and get world class experience with this products. Register here. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, SQLAuthority News, T SQL, Technology Tagged: SQL PASS

    Read the article

  • SQL SERVER – 2008 – Introduction to Snapshot Database – Restore From Snapshot

    - by pinaldave
    Snapshot database is one of the most interesting concepts that I have used at some places recently. Here is a quick definition of the subject from Book On Line: A Database Snapshot is a read-only, static view of a database (the source database). Multiple snapshots can exist on a source database and can always reside on the same server instance as the database. Each database snapshot is consistent, in terms of transactions, with the source database as of the moment of the snapshot’s creation. A snapshot persists until it is explicitly dropped by the database owner. If you do not know how Snapshot database work, here is a quick note on the subject. However, please refer to the official description on Book-on-Line for accuracy. Snapshot database is a read-only database created from an original database called the “source database”. This database operates at page level. When Snapshot database is created, it is produced on sparse files; in fact, it does not occupy any space (or occupies very little space) in the Operating System. When any data page is modified in the source database, that data page is copied to Snapshot database, making the sparse file size increases. When an unmodified data page is read in the Snapshot database, it actually reads the pages of the original database. In other words, the changes that happen in the source database are reflected in the Snapshot database. Let us see a simple example of Snapshot. In the following exercise, we will do a few operations. Please note that this script is for demo purposes only- there are a few considerations of CPU, DISK I/O and memory, which will be discussed in the future posts. Create Snapshot Delete Data from Original DB Restore Data from Snapshot First, let us create the first Snapshot database and observe the sparse file details. USE master GO -- Create Regular Database CREATE DATABASE RegularDB GO USE RegularDB GO -- Populate Regular Database with Sample Table CREATE TABLE FirstTable (ID INT, Value VARCHAR(10)) INSERT INTO FirstTable VALUES(1, 'First'); INSERT INTO FirstTable VALUES(2, 'Second'); INSERT INTO FirstTable VALUES(3, 'Third'); GO -- Create Snapshot Database CREATE DATABASE SnapshotDB ON (Name ='RegularDB', FileName='c:\SSDB.ss1') AS SNAPSHOT OF RegularDB; GO -- Select from Regular and Snapshot Database SELECT * FROM RegularDB.dbo.FirstTable; SELECT * FROM SnapshotDB.dbo.FirstTable; GO Now let us see the resultset for the same. Now let us do delete something from the Original DB and check the same details we checked before. -- Delete from Regular Database DELETE FROM RegularDB.dbo.FirstTable; GO -- Select from Regular and Snapshot Database SELECT * FROM RegularDB.dbo.FirstTable; SELECT * FROM SnapshotDB.dbo.FirstTable; GO When we check the details of sparse file created by Snapshot database, we will find some interesting details. The details of Regular DB remain the same. It clearly shows that when we delete data from Regular/Source DB, it copies the data pages to Snapshot database. This is the reason why the size of the snapshot DB is increased. Now let us take this small exercise to  the next level and restore our deleted data from Snapshot DB to Original Source DB. -- Restore Data from Snapshot Database USE master GO RESTORE DATABASE RegularDB FROM DATABASE_SNAPSHOT = 'SnapshotDB'; GO -- Select from Regular and Snapshot Database SELECT * FROM RegularDB.dbo.FirstTable; SELECT * FROM SnapshotDB.dbo.FirstTable; GO -- Clean up DROP DATABASE [SnapshotDB]; DROP DATABASE [RegularDB]; GO Now let us check the details of the select statement and we can see that we are successful able to restore the database from Snapshot Database. We can clearly see that this is a very useful feature in case you would encounter a good business that needs it. I would like to request the readers to suggest more details if they are using this feature in their business. Also, let me know if you think it can be potentially used to achieve any tasks. Complete Script of the afore- mentioned operation for easy reference is as follows: USE master GO -- Create Regular Database CREATE DATABASE RegularDB GO USE RegularDB GO -- Populate Regular Database with Sample Table CREATE TABLE FirstTable (ID INT, Value VARCHAR(10)) INSERT INTO FirstTable VALUES(1, 'First'); INSERT INTO FirstTable VALUES(2, 'Second'); INSERT INTO FirstTable VALUES(3, 'Third'); GO -- Create Snapshot Database CREATE DATABASE SnapshotDB ON (Name ='RegularDB', FileName='c:\SSDB.ss1') AS SNAPSHOT OF RegularDB; GO -- Select from Regular and Snapshot Database SELECT * FROM RegularDB.dbo.FirstTable; SELECT * FROM SnapshotDB.dbo.FirstTable; GO -- Delete from Regular Database DELETE FROM RegularDB.dbo.FirstTable; GO -- Select from Regular and Snapshot Database SELECT * FROM RegularDB.dbo.FirstTable; SELECT * FROM SnapshotDB.dbo.FirstTable; GO -- Restore Data from Snapshot Database USE master GO RESTORE DATABASE RegularDB FROM DATABASE_SNAPSHOT = 'SnapshotDB'; GO -- Select from Regular and Snapshot Database SELECT * FROM RegularDB.dbo.FirstTable; SELECT * FROM SnapshotDB.dbo.FirstTable; GO -- Clean up DROP DATABASE [SnapshotDB]; DROP DATABASE [RegularDB]; GO Reference : Pinal Dave (http://blog.SQLAuthority.com) Filed under: SQL, SQL Authority, SQL Backup and Restore, SQL Data Storage, SQL Query, SQL Server, SQL Tips and Tricks, SQLServer, T SQL, Technology

    Read the article

  • SQL SERVER – 2012 – Summary of All the Analytic Functions – MSDN and SQLAuthority

    - by pinaldave
    SQL Server 2012 (RC0 Available here) has introduced new analytic functions. These functions were long awaited and I am glad that they are here. Previously when any of this function was needed people use to write long T-SQL code to simulate that and now no need of the same. Having available native function also helps performance as well readability. In last few days I have written many articles on this subject on my blog. The goal was make these complex analytic functions easy to understand and make it widely accepted. As this new functions are available and as awareness spreads we should start using the new functions. Here is the quick list of the new function and relevant MSDN site. Function SQLAuthority MSDN CUME_DIST CUME_DIST CUME_DIST FIRST_VALUE FIRST_VALUE FIRST_VALUE LAST_VALUE LAST_VALUE LAST_VALUE LEAD LEAD LEAD LAG LAG LAG PERCENTILE_CONT PERCENTILE_CONT PERCENTILE_CONT PERCENTILE_DISC PERCENTILE_DISC PERCENTILE_DISC PERCENT_RANK PERCENT_RANK PERCENT_RANK I also enjoyed three different puzzles during the course of this series which gave clear idea to the SQL Server 2012 analytic functions. SQL SERVER – Puzzle to Win Print Book – Functions FIRST_VALUE and LAST_VALUE with OVER clause and ORDER BY SQL SERVER – Puzzle to Win Print Book – Write T-SQL Self Join Without Using LEAD and LAG SQL SERVER – Puzzle to Win Print Book – Explain Value of PERCENTILE_CONT() Using Simple Example This series will be always my dear series as during this series I had went through very unique experience of my book going out of stock and becoming available after 48 hours. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Function, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

< Previous Page | 9 10 11 12 13 14 15 16 17 18 19 20  | Next Page >