Database creation - PowerPoint PPT Presentation

1 / 31
About This Presentation
Title:

Database creation

Description:

Database creation Using a script: Connect internal as sysdba; create database NAME .; Using the OEM GUI Parameters stored in a small DB Access to other utility pgms ... – PowerPoint PPT presentation

Number of Views:98
Avg rating:3.0/5.0
Slides: 32
Provided by: Frede168
Category:

less

Transcript and Presenter's Notes

Title: Database creation


1
Database creation
  • Using a script
  • Connect internal as sysdba
  • create database NAME.
  • Using the OEM GUI
  • Parameters stored in a small DB
  • Access to other utility pgms eg RMAN
  • Can accommodate several DBAs
  • Key decisions on sizing
  • Eg assign a block size once and for all

2
Block size
  • Smallest unit of storage in oracle DB
  • Critical perf. Factor as inefficient space use
    leads to increase I/O activity
  • Block overhead
  • Integrity section (90 bytes)
  • Runtime overhead (136 bytes)
  • Transactional header (24 bytes and growing)
  • Directory section (Pointers)
  • And then..data
  • Big is beautiful? Percentage of overhead
    decreases when block size increases, but no right
    answer here


Approx. 256 b
3
2 K, 4K, 8K or 16K?
  • OLTP prefer smaller small granularity increase
    likelihood that data is in buffer (indexed
    access)
  • DW or DSS prefer larger physically close data
    used together less movement
  • In practice constraint on memory will also
    dictate
  • Buffer size now split in 3 areas
  • KEEP pool
  • RECYCLE pool
  • DEFAULT pool

4
Block size and row size
  • Table with average row length 2236 b (10m
    records)
  • Block size 4k 4006 b gt one row per block
  • 8 k gt 3 rows per block

5
Opening and closing the DB
  • Startup open NAME
  • Or
  • Startup mount NAME
  • Alter database open
  • Shutdown
  • Else use OEM menus

6
When DB is started
page846
  • System Global Area (SGA) created
  • Data block buffer cache created
  • Size is key for performance of DB
  • fraction of whole DB
  • DB_BLOCK_BUFFERS / DB_BLOCK_SIZE
  • Least Recently Used (LRU) method
  • Too small means fight for space between objects
    gt increase in misses gt increase in I/O
  • Too big crippling for other operations

7
When DB created (2)
  • Shared SQL Pool created
  • Stores DD cache info about data structure
  • Stores library cache info about SQL statements
    run
  • LRU algorithm
  • Too small gt no re-use and low performance
  • SHARED_POOL_SIZE in init.ora

8
Parameter files INIT.ORA
  • Initialisation file for each DB
  • MYDB gt initmydb.ora
  • Contains sizing of SGA and number of parameters,
    DB name and block size
  • Read at startup means changes only implemented at
    next startup

9
The control file
  • Each DB also has a control file
  • Contains Dbname, name and location of files and
    redo logs, timestamp of DB creation
  • Can be mirrored for added security
  • Identifies all files that must be opened when Db
    is loaded
  • Also used in DB recovery

10
Data hierarchy in Oracle
  • Data hierarchy in Oracle is specific (and
    demented!)
  • Normally, unit of storage is file
  • Managed by OS
  • Can grow / shrink
  • Physical sections logically connected by OS
  • In Oracle a DB one or more tablespaces
  • One or more objects, indexes, clusters
  • Fixed size gets full
  • Extended by DBA only
  • One file can support only ONE tablespace

See figure 3.1
11
About tablespaces
  • Kind of partitioning
  • Except fixed size
  • Extended by dba
  • Oracle very strong at TS level
  • Optimal seems to be around 2GB
  • Strategy isolate aged data into one tablespace
    gt can be made read-only
  • Should store similar objects together

12
Reasoning behind TS creation
  • Business relevance
  • Size of objects
  • Type of activity
  • Volume of activity
  • Backup
  • Transportability
  • Time-based changes in activity

13
Reasoning behind TS creation
P 148
  • Business relevance
  • Size of objects
  • Type of activity
  • Volume of activity
  • Backup
  • Transportability
  • Time-based changes in activity

e.g. modular design / keep data togegther
Object of similar sizes reuse of space
e.g. index VS table / full scan VS rowID
Low I/O VS high I/O objects
TS is smallest unit for Bcup and restore
Cloning copy is done by TS as well
DW env. Aged archived system better access time
related
Bias towards more rather than less gt
compromise Also perf reading
14
Create a tablespace
  • Create tablespace FREDDY datafile
  • path\freddy.dbf size 1000K
  • Default storage (initial 10K next 10K
  • Minextent 1 maxextents 100
  • Pctincrease 0)
  • Permanent
  • 10 10 10 10 10 10 10 10 10 10
  • 20 30 40 50 60 70 80 90 100
  • 10 10 12 14.4 17.2 20.7 25 30 36 42.3
  • 10 20 32 46.4 63.6 84.3 109.3 139.3 175.3 218.5

15
Object creation
  • Tables one physical space on disk segment
    same for any object
  • Each segment space set out in tablespace
    initial extent
  • When table grows allocate next extent set of
    adjacent blocks
  • Until tablespace is full
  • see figure 20.3

16
What a segment looks like
  • Each segment contains a header block
  • Space directory for other blocks
  • Properly sized tables few extents (low
    fragmentation)
  • Additional extents are never lost until drop or
    delete commands are used
  • Can also adjust pctfree value (to fill a Null
    value column)

17
Sizing extents
  • Try to match the max capacity of your storage
    device to limit I/O operations
  • For single objects, It is beneficial to keep
    extents of same size (deletion)
  • Try to limit the number of extents below 50
    (measure of the growth of the object)
  • If too many, rebuild object with larger size
    extents

18
Oracle space allocation
  • DBA allocates space for table after 2 days of
    hard thinking
  • Oracle ignores it blatantly as follows

Table name initial size next extent oracle
sizing next extent Smalltab 7K 7K 20K 20K Medi
umtab 103K 103K 120K 120K Total size after
growth smalltab 40K instead of 14K mediumtab
206K instead of 240K
19
Creating and using tables
  • Table creation with CREATE TABLE statement
  • Syntax
  • Create table NAME (
  • field1 datatype special constraint
  • field2 datatype etc
  • )
  • Special constraint e.g. NOT NULL or PRIMARY KEY
    or CHECK

20
Constraints
  • Candidate key
  • Create table fred (
  • name varchar2(10) NOT NULL
  • firstname varchar2(10) NOT NULL
  • age number
  • Constraint FRED_UQ unique (name, firstname)
  • )

21
Primary Key
  • Create table fred (
  • name varchar2(10)
  • firstname varchar2(10)
  • age number
  • Constraint FRED_PK Primary Key (name, firstname)
  • )
  • Same apart always NOT NULL can only have one
    of them in a table
  • If only one field just list primary key as
    special constraint

22
Foreign key aka referential integrity constraint
  • Create table fred (
  • ID number NOT NULL
  • name varchar2(10)
  • firstname varchar2(10)
  • age number
  • Constraint FRED_PK Primary Key (ID), foreign key
    (age) references OTHERTAB(age)
  • )
  • Use ON DELETE CASCADE to maintain constraint

23
Check constraint
  • Create table fred (
  • ID number NOT NULL
  • name varchar2(10)
  • firstname varchar2(10)
  • age number CHECK (age between 18 and 65)
  • )

24
Deleting data
  • Drop table statement remove table
  • Or truncate table statement empty it
  • Warning neither can be rolled back
  • Otherwise
  • Alter table NAME Add ()
  • Alter table NAME modify ()
  • drop column fieldname
  • See rules on handout

25
Now what? Clusters
  • To store different tables in physically close
    areas on disk
  • Create a cluster to store both objects
  • Cluster key is any useful field that links the
    tables
  • Can keep name of field or pick new one
  • e.g. Create cluster fred (name datatype)
  • then assign tables create tablecluster fred
    (name)

26
What this cluster looks like
  • 12 Cork fred adam French
  • Midleton Bob Fitz Irish
  • Bally Brian Oreilly English

Table 1
Table 2
Cluster Key
27
Finally
  • Users can have different logical views on a
    database
  • These views are called..?
  • They allow for a re-organisation of physical DB
    objects for each users
  • Each ? can contain elements from different
    tablespaces
  • Has limited size allocated to each user

28
Summary storage in Oracle
  • Keywords tablespace / datafiles / database
    objects / segments / extents
  • When you create an object
  • Specify size / tablespace info storage clause
  • Otherwise default values
  • Pctincrease should be used with care (affects
    re-usability of reclaimed space i.e. extents)
  • Wise use of tablespace eases admin (like
    partitioning)
  • sizing follows sizing of device e.g. 2GB
  • Many utilities to work with tablespaces
  • Isolate processes e.g. a report running inside
    a TS
  • Also for archiving

29
Performance issues
  • Extent sizing a key
  • Oracle has 2 ways to read data
  • By rowID typically thru an index
  • Full table scan
  • In second case, oracle reads several blocks at a
    time
  • Set with DB_FILE_MULTIBLOCK_READ_COUNT
  • Limited only by I/O device buffer capacity
  • E.g. block 4K, I/O buffer 64K gt ?????
  • Extent size should be multiple of 64K

30
example
  • Table with 10 extents 64 K each
  • I/O device buffer 64K gt 10 reads
  • If extent 640K gt ???? Reads
  • If 8 extents 80K each gt ??? Reads
  • Reads cannot span extents gt 16 reads or an
    increase of 60 in I/O operations
  • Cl either use extents much larger than buffer or
    use multiple of buffer size

31
And for roll back segments
  • Should be kept small enough
  • Current wisdom is to ensure TS for rollback is
    normally 50 free
  • Not always possible to use only one RB size
  • For different types of processes e.g. TPS
  • Use two different TB to isolate activites
  • Easier trouble shooting
  • Discarded rollback segments go into the log file
Write a Comment
User Comments (0)
About PowerShow.com