SQL Server Index Basics


Non Clustered and Clustered Index:

The primary key created for the StudId column will create a clustered index for the Studid column. A table can have only one clustered index on it. When creating the clustered index, SQL server 2005 reads the Studid column and forms a Binary tree on it. This binary tree information is then stored separately in the disc.

A non-clustered index is useful for columns that have some repeated values. Say for example, AccountType column of a bank database may have 10 million rows. But, the distinct values of account type may be 10-15. A clustered index is automatically created when we create the primary key for the table. We need to take care of the creation of the non-clustered index.

Recommendation:
-Indexes with INCLUDEs are a big win for frequently-called range queries.
-If access is always per-record on the clustered index, don’t add indexes with INCUDEs.
-For frequently called relatively-static data, consider using multiple indexes with INCLUDEs to cover different queries even if the KeyColumns are the same.
-Where you have created covering indexes where all the columns are in the KeyColumns, consider changing these indexes to have the minimal set of KeyColumns and the other columns in the IncludedColumns. This will lower the space requirements and speed the retrieval of records.


Ref:
http://www.codeproject.com/Articles/173275/Clustered-and-Non-Clustered-Index-in-SQL-2005
http://msdn.microsoft.com/en-us/library/ms188783.aspx
http://www.databasejournal.com/features/mssql/article.php/3787021/Exploring-SQL-Servers-Index-INCLUDEs.htm
sql/article.php/3787021/Exploring-SQL-Servers-Index-INCLUDEs.htm

Comments

Post a Comment

Popular posts from this blog

Restoring Master, Model and MSDB from Netbackup

Restoring System Databases got two parts: Taking the instance in single user mode and Restoring the Master Database If you need to rebuild the system databases on top of this you can use the below command: "C:\Program Files\Microsoft SQL Server\ 100 \Setup Bootstrap\ SQLServer2008R2 \setup.exe" /ACTION=REBUILDDATABASE /INSTANCENAME=SQL008 /SQLSYSADMINACCOUNTS=domain\ServiceUserName /SAPWD="yourpasswordOfSAuser" /SQLCOLLATION="Latin1_General_CI_AS" here 100 is for SQL Server 2008R2. It can vary case by case. How to take instance to single user mode: Two situation can happen: 1. Standalone Instance  2. Failover Cluster Instance 1. Standalone Instance: Go to SQL Config Manager SQL Service Properties -> Advanced Properties -> Startup Parameters put -m; in the Startup Parameters Apply and Okay - it will ask to restart the SQL Service to take effect. Stop and Start the service from Config Manager or you can stop and serv...
Read more

Oracle Job Operations

As a DBA, often I need to deal with Client's Jobs. Sometimes the jobs got hault, needs to break etc etc. Here are some useful commands for this sort of operations. ## Job Owner Schema can do this execute dbms_job.remove(444); execute dbms_job.broken(444, true); ## If Sys User needs to remove the job exec SYS.DBMS_IJOB.REMOVE(444); ## to find which jobs are running SELECT a.sid, c.serial#, a.job, a.failures, to_char(a.this_date, 'mm/dd/yyyy hh:mi pm') startdatetime, b.what FROM dba_jobs_running a, dba_jobs b, v$session c WHERE a.job = b.job AND a.sid = c.sid order by a.this_date ##let say we have only sid and serial and we need to kill the process from both OS and database. So first we need to find out the SPID. this query for any RAC database select P.SPID ,S.SID ,S.SERIAL#, S.SCHEMANAME, S.PROGRAM, S.INST_ID, S.OSUSER, S.STATUS from gv$process p, gv$session s where P.ADDR=S.PADDR and S.SID=888 and s.serial#=9293; ## kill the job al...
Read more