Enterprise dbs and Database indexing

ENTERPRISE DB SERVER AN INTRODUCTION
Why do we need to know structure, features of an enterprise database?
As Batman said: Either you die yourself a hero or live long enough to see yourself becoming a villain.
(The Dark Knight)
Microsoft, Oracle & even Apple: Lived long enough to see themselves becoming villains. Web does
not like them, open source community dislikes them. :)
(Antitrust, proprietary code, patent war, aggressive sales tactics, Market Monopoly, DRM)
But Microsoft Windows Server Software still resides on 16% web server machines, 54% of all
fortune 1000 company servers. Not to mention many corporate SME use entire microsoft stack in
their intranet making it top intranet based software.
War is Drug: The Hurt Locker.
We can say Microsoft stack is Drug. Once you are into it You
can come out of it.
my precious code…

Why?
Microsoft SQL Server may not be the best enterpriseDB out there but it has the best Software
Suit in terms of packaging of various components, features and analytics verticals.
Basic components of the sql server database is as follows
Sql Server Database engine (DB Objects, Jobs, Query Tools, Tuning tools, Full-Text, Log and DTC)
Sql Server Analysis Services (OLAP, Cubes, PostGre In Memory, BI, Business Analytics)
Sql Server Reporting engine (Charting, Matrix, Templates, Sub Reports)
Sql Server Integration services (DTS Packages)
[1] SQL Server Database Engine: This part of SQL Server actually creates and drives relational
databases. (profiler, Agent, DB Mail, Browser, Pipes, mdf and ldf, Recovery, Mirror, RAIDs)
[2] SQL Server Analysis Services (SSAS): SSAS is the data-analysis component of SQL Server. It
can create OLAP (OnLine Analytical Processing) cubes sophisticated programming objects for
organizing data inside a relational database and do data mining (pulling relevant data out of a
database in response to an ad-hoc question).
[3] SQL Server Reporting Services (SSRS): SSRS is a component of SQL Server that provides
reporting regardless of a database’s operating system.
[4] SQL Server Integration Services (SSIS): SSIS is a component of SQL Server that does the
Extract, Transform, and Load (ETL) process that cleans up and formats raw data from source
systems for inclusion in the database as ready-to-use information.

SQL
SERVER
MY SQL
Feature Comparison

DB INDEXING
•Extent
–8 contiguous 64KByte size data Pages
–Once extent full, next record will take up a whole additional extent.
–Pre-allocating space saves time.
•Page
–64K bytes
–# of records/page varies with bytes/record
–Types of Pages: Data and Index pages
–Page Split
–When page becomes full, it splits
–New page allocated: ½ data from old page moved to new (FILL FACTOR*)
•Rows
–Storage is different as per different versions
–Text data or Blob data is stored differently, only finite data is stored in Pages

Extent
Page 1 Page 2 ……….
Page 8
TABLE

FRAGMENTATION / FILL FACTOR / SPARSE COLUMNS
Index PAGE
Amit - Page 1
bharat - Page 1
Chintu - Page 1
--40--
Index PAGE
Jay - Page N
Mohan - Page N
Rajesh - Page N
--40--
Index PAGE
Sumit - Page M
Siddu - Page M
Zakir - Page M
---40---
TABLE
No of Records
120
Index PAGE
Amit - Page 1
bharat - Page 1
Chintu - Page 1
--40--
Index PAGE
Jay - Page N
Mohan - Page N
Rajesh - Page N
--40--
Index PAGE
Sumit - Page M
Siddu - Page M
Zakir - Page M
---40---
Index PAGE
Chinu - Page X

TABLE
No of Records
121
Index PAGE
Amit - Page 1
bharat - Page 1
Chintu - Page 1
Chinu - Page X
----empty----
Index PAGE
Jay - Page N
Mohan - Page N
Rajesh - Page N
------empty-----
------empty-----
Index PAGE
Sumit - Page M
Siddu - Page M
Zakir - Page M
------empty-----
------empty-----
SPARSE COLUMNS
Index PAGE
Amit,24 - Page 1
bharat,22 - Page 1
Chintu,43 - Page 1
Chinu,45 - Page X

● Dense Index and sparse Index
Dense for all search key values
Sparse for few search key values
● Sparse Index:
-Less space and less maintenance overhead for insertions and deletions.
-Generally slower than dense index for locating records.
-sparse index with an index entry for every block in file, corresponding to least
search-key value in the block.
- only when data is ordered (sometimes internally maintained by database)
- Find index record with largest search-key value < K, then start searching reverse

Points:
● Updating indices imposes overhead on database modification --when a
row is modified, every index on the row must be updated. e.g. Log Table
● High Insert values: Bulk insert, BCP routines
● Sequential scan using clustered index is efficient, but a sequential scan
using a non clustered index is expensive.
● Only one clustered index, hundreds of non clustered index (ascending,
descending)
● Where > clause = clustered index
● Where = clause Non clustered index
● Index Maintenance is also important due to fragmentation
● Full Text

Choosing between clustered and non-clustered index:
Five Factors involved when choosing the indexing technique
• Access type - is the type of access being used.
• Access time - time required to locate the data.
• Insertion time - time required to insert the new data.
• Deletion time - time required to delete the data.
• Space overhead - the additional space occupied by the added data
structure.

Hopefully ☺
● Database Components & features
● Sql Constructs
● Database design & Normalization
● Analytics & Reporting Concepts
● Database Maintenance & tuning
● Database Architectures

Enterprise dbs and Database indexing

More Related Content

Viewers also liked (6)

Similar to Enterprise dbs and Database indexing (20)

Recently uploaded (20)

Enterprise dbs and Database indexing