1
© The Pythian Group Inc., 2018
Aishwarya Kala
Leveraging Partitioning
Enhancements
2
© The Pythian Group Inc., 2018 2
© The Pythian Group Inc., 2019
ABOUT PYTHIAN
Pythian’s 400+ IT professionals help
companies adopt and manage
disruptive technologies to better
compete
3
© The Pythian Group Inc., 2018
AGENDA
3
© The Pythian Group Inc., 2019
● Introduction
● Partitioning 101
● Partition Extensions + New Features
● Partition Manageability + New Features
● ADO & Data Archival
● Some More New Features
4
© The Pythian Group Inc., 2018
Partitioning 101
5
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 5
● Divide and Conquer
● Divides Objects like tables, indexes, index-organized tables into smaller, more manageable
pieces, called partitions.
● Manage partitions individually as well as collectively.
● The decision to assign a row to a partition is taken based on a Partitioning Key ( one or more
columns of the table)
Partitioning
6
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 6
Partitioning
01-Jan-19
02-Jan-19
03-Jan-19 03-Feb-19
02-Feb-19
01-Feb-19
03-Mar-19
02-Mar-19
01-Mar-19
01-Jan-19
02-Jan-19
03-Jan-19
01-Feb-19
02-Feb-19
03-Feb-19 03-Mar-19
02-Mar-19
01-Mar-19
Table : Orders
Jan19 Mar19
Feb19
Table : Orders
Data distributed by
order_date
7
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 7
● Standard Data Distribution Methods
● Single-Level Partitioning [RANGE | LIST | HASH ]
● Composite Partitioning - subpartitions [RANGE | LIST | HASH ] || [RANGE | LIST | HASH ]
Partitioning Strategies
Jan19
Feb19
Mar19
RANGE
ASIA
EMEA
AFRICA
HASH
LIST
H1
H2
H3
H4
8
© The Pythian Group Inc., 2018
Partition Extensions + New Features
9
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 9
● Extension of Range Based Partitions (Introduced in 11g)
● Auto creates partitions when data exceeds all the existing ranges
● Subtle Differences
● Interval Based partitions do not have Max-value Partition
● Naming & Tablespace Constraints
Interval Partitioning
10
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 10
● Extension of Range Based Partitions (Introduced in 11g)
● Auto creates partitions when data exceeds all the existing ranges
● Subtle Differences
● Interval Based partitions do not have Maxvalue Partition
● Naming & Tablespace Constraints
Interval Partitioning
MAXVAULE
Range Based Partitioning
Jan19 Feb19 Jan19
New Record with
order_date of
02-Mar-2019
Interval Based Partitioning
11
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 11
● Extension of Range Based Partitions (Introduced in 11g)
● Auto creates partitions when data exceeds all the existing ranges
● Subtle Differences
● Interval Based partitions do not have Maxvalue Partition
● Naming & Tablespace Constraints
● Range Based partitions do not have any Lower Bounds
Interval Partitioning
MAXVAULE
Range Based Partitioning
Interval Based Partitioning
Jan19 Feb19 Jan19
SYS_NAME
New Record with
order_date of
02-Mar-2019
12
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 12
Partition Bounds
MAXVAULE
Range Based Partitioning
Values less than Feb 2019
Values less than Apr 2019
Values belonging to Jan 2019
Values belonging to Apr 2019
Interval Based Partitioning
New Record with
order_date of
02-Mar-2019
13
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 13
● Extension of Range Based Partitions (Introduced in 11g)
● Auto creates partitions when data exceeds all the existing ranges
● Subtle Differences
● Interval Based partitions do not have Maxvalue Partition
● Naming & Tablespace Constraints
● Range Based partitions do not have any Lower Bounds
Be wary of jobs that do not query data dictionary views for partition names, when moving from
range to interval partitioning.
Interval Partitioning
14
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 14
● Automatic List Partitions
● Extension of List Based Partitions ( Introduced in 12.2 )
● Auto creates partitions when no corresponding partition key value found.
● The decision is to choose between
o List + Default
o Auto List
● Multi-Column List Partitioning ( Introduced in 12.2 )
Only one Default bucket
List Partitioning Enhancements
Not Suitable when too many distinct
values expected.
15
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 15
● Automatic List Partitions
● Extension of List Based Partitions ( Introduced in 12.2 )
● Auto creates partitions when no corresponding partition key value found.
● Not Suitable when too many distinct values expected.
● The decision is to choose between
o List + Default
o Auto List
● Multi-Column List Partitioning ( Introduced in 12.2 )
Only one Default bucket
List Partitioning Enhancements
D
E
F
A
U
L
T
ASIA ASIA EMEA
INDIA NEPAL U.K.
16
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 16
● Referential Partitioning
● Child table inherits the partitioning strategy of parent table through Referential Constraints
● Reduces Data Maintenance Overhead
■ A split operation on Parent will split the dependent child partitions.
■ Dropping a partition in parent table, will drop the dependent child partition – No need to disable the FK-PK
■ Applies to Truncate as well. (TRUNCATE CASCADE)
● Virtual Column Based Partitioning
● Partition based on the virtual columns of a table.
Other Extensions
17
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 17
● What do you want to achieve ?
● Performance
 Partition Pruning
 Partition-wise join
 DML Speed up
 Data Load
● Data Management Benefits
● Do you know your data ?
● Type of Data - Records, transactions…
● Data Access Patterns
● Data Lifecycle Requirements.
Choosing a Partitioning- Strategy
18
© The Pythian Group Inc., 2018
Partition Manageability +
New Features
19
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 19
● Partition a Stand-Alone Table
1. Create a separate Partitioned Table + Transfer Data
2. Exchange Partitioning
■ Create Separate Tables with Filtered Data
■ Transfer the Filtered Data into a corresponding partition
3. DBMS_REDEFINITION
Online Conversion of a Non-Partitioned to a Partitioned Table
CREATE TABLE ORDERS_2018
where Order_DATE < 2018 ;
Alter table ORDERS_PART Exchange
partition ORDERS_PART2018 with table
orders_2018;
20
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 20
Online Conversion of a Non-Partitioned to a Partitioned Table
21
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 21
● Global partitioned
● Non-prefixed indexes
● Prefixed indexes
● Bitmap indexes
Online Conversion of a Non-Partitioned to a Partitioned Table
No Effect
Global Non-partitioned Indexes
Local Partitioned Indexes
Local Partitioned Indexes
Index that contains the
leading column of the
partitioning key
Partitioning_key =>
order_date
Index => (order_Date,
region)
22
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 22
Online Split & Move Operations
23
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 23
Online Split & Move Operations
• Indexes always remain ONLINE
• Concurrent Query & DML Operations allowed – not preferred => Impacts compression ratio
24
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 24
Create Table For Exchange
TAB
ORA-14097: column
type or size
mismatch …
CREATE TABLE TAB
FOR EXCHANGE WITH TABLE PART_TAB;
PART_TAB
25
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 25
Data Management
Jan19
Feb19
Mar19
Table Size 300 GB
Table Details Monthly Range-based Partitions
Requirements
• Compress All Data older than 3 months (~25 GB/Month)
• Move data older than a year from SSD To HDD
26
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 26
Data Management
Jan19
Feb19
Mar19
Requirements
• Compress All Data older than
3 months (~25 GB/Month)
• Move data older than a year
from SSD To HDD
• ALTER TABLE ORDER_HISTORY MOVE PARTITION OH_MAY2019 COMPRESS ;
• ALTER TABLE ORDER_HISTORY MODIFY PARTITION OH_MAY2019 REBUILD
UNUSABLE LOCAL INDEXES;
Compression Steps
• ALTER TABLE ORDER_HISTORY MOVE PARTITION OH_JUN2018 TABLESPACE ORDER_HISTORY_SLOW ;
• ALTER TABLE ORDER_HISTORY MODIFY PARTITION OH_JUN2018 REBUILD UNUSABLE LOCAL INDEXES
TABLESPACE ORDER_HISTORY_SLOW ;
Data Archiving Steps
27
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 27
Data Management
Manual Operation Involved
Repetitive - Every Month
Multiple Partitioned Tables
28
© The Pythian Group Inc., 2018
Automatic Data
Optimization (ADO)
29
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 29
Heat Maps
30
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 30
Heat Maps
● Track Data Access
● Reads vs Writes
● Types of Reads – FTS vs Index Lookups
● Row, Segment, Tablespace Level
● Access the information via Data Dictionary Views
● Database-aware
● Maintenance jobs, stats gathering, backups, don’t affect heat map.
● Doesn’t track objects in SYSTEM and SYSAUX tablespaces
HEAT_MAP = ON
Advanced Compression
License
31
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 31
Heat Maps
32
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 32
Heat Maps
33
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 33
Heat Maps
34
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 34
Heat Maps
35
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 35
Heat Maps
36
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 36
Automatic Data Optimization (ADO)
● Policies that Manage Data Automatically
Compression
Data
Movement
HOW ??
37
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 37
Automatic Data Optimization (ADO)
● Policies that Manage Data Automatically
Compression
Data
Movement
HOW ??
Moving to tablespaces on
separate storage tiers.
38
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 38
Automatic Data Optimization (ADO)
● Policies that Manage Data Automatically
Compression
Data
Movement
HOW ??
Data can be compressed at
Row storage level
Segment level
Group (multiple objects) level
39
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 39
Automatic Data Optimization (ADO)
● Policies can be
● Space Based
 E.g. When tablespace utilization reaches 85%, then move to another tablespace
● Time And Access method Based
 E.g. Compress any row that has not been updated for 90
 E.g. Move a segment that has not been updated for 180 to another tablespace
TRIGGERS
40
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 40
Automatic Data Optimization (ADO) – Time & Access Method
Condition of the Policy
41
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 41
Automatic Data Optimization (ADO) – Time & Access Method
Goal of the Policy
42
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 42
Automatic Data Optimization (ADO) – Time & Access Method
Scope of the Policy
43
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 43
Automatic Data Optimization (ADO) – Time & Access Method
44
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 44
Automatic Data Optimization (ADO) – Space
Goal of the Policy
Scope of the Policy
Condition of the Policy???
45
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 45
Automatic Data Optimization (ADO) – Space
Condition of the Policy???
• Global Defaults
• Can be controlled
dbms_ilm_admin.customize_ilm
46
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 46
Automatic Data Optimization (ADO) - Execution
● Row Level
● Remaining Policies
● Supports Manual Evaluation & Execution
Setup custom jobs for larger tables that evaluate the policy periodically
MMON (15 mins)
Maintenance Window
47
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 47
Manual Execution
48
© The Pythian Group Inc., 2018
Some More New Features
49
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 49
● Protect Data from accidental modifications
● Columns can be added /modified
● RO Partitions Cannot be Dropped
● Benefits of RO Partitions vs RO Tablespaces
● No Need for separate tablespaces
● Benefits smaller partitions
Read Only Partitions
ALTER TABLE ORDERS MODIFY PARTITION
ORDERS_2018 READ ONLY;
50
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 50
● Applicable to Partition Move, Split, Merge
● Only rows matching the filter condition are retained
Filtered Partition Maintenance Operations
51
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 51
● Partition External Tables
● Combines of Regular Tables + External Partitioned Tables
Hybrid Partitioned Tables
52
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 52
● Change the partitioning strategy
● Avoids Manual Rebuild
● Uses the MODIFY Clause
Modifying Partitioning Strategy
53
© The Pythian Group Inc., 2018
© The Pythian Group Inc., 2019 53
Modifying Partitioning Strategy
54
© The Pythian Group Inc., 2018 54
© The Pythian Group Inc., 2019
THANK YOU
You can reach me :
@aishwaryakala13

Leveraging partition enhancements

  • 1.
    1 © The PythianGroup Inc., 2018 Aishwarya Kala Leveraging Partitioning Enhancements
  • 2.
    2 © The PythianGroup Inc., 2018 2 © The Pythian Group Inc., 2019 ABOUT PYTHIAN Pythian’s 400+ IT professionals help companies adopt and manage disruptive technologies to better compete
  • 3.
    3 © The PythianGroup Inc., 2018 AGENDA 3 © The Pythian Group Inc., 2019 ● Introduction ● Partitioning 101 ● Partition Extensions + New Features ● Partition Manageability + New Features ● ADO & Data Archival ● Some More New Features
  • 4.
    4 © The PythianGroup Inc., 2018 Partitioning 101
  • 5.
    5 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 5 ● Divide and Conquer ● Divides Objects like tables, indexes, index-organized tables into smaller, more manageable pieces, called partitions. ● Manage partitions individually as well as collectively. ● The decision to assign a row to a partition is taken based on a Partitioning Key ( one or more columns of the table) Partitioning
  • 6.
    6 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 6 Partitioning 01-Jan-19 02-Jan-19 03-Jan-19 03-Feb-19 02-Feb-19 01-Feb-19 03-Mar-19 02-Mar-19 01-Mar-19 01-Jan-19 02-Jan-19 03-Jan-19 01-Feb-19 02-Feb-19 03-Feb-19 03-Mar-19 02-Mar-19 01-Mar-19 Table : Orders Jan19 Mar19 Feb19 Table : Orders Data distributed by order_date
  • 7.
    7 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 7 ● Standard Data Distribution Methods ● Single-Level Partitioning [RANGE | LIST | HASH ] ● Composite Partitioning - subpartitions [RANGE | LIST | HASH ] || [RANGE | LIST | HASH ] Partitioning Strategies Jan19 Feb19 Mar19 RANGE ASIA EMEA AFRICA HASH LIST H1 H2 H3 H4
  • 8.
    8 © The PythianGroup Inc., 2018 Partition Extensions + New Features
  • 9.
    9 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 9 ● Extension of Range Based Partitions (Introduced in 11g) ● Auto creates partitions when data exceeds all the existing ranges ● Subtle Differences ● Interval Based partitions do not have Max-value Partition ● Naming & Tablespace Constraints Interval Partitioning
  • 10.
    10 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 10 ● Extension of Range Based Partitions (Introduced in 11g) ● Auto creates partitions when data exceeds all the existing ranges ● Subtle Differences ● Interval Based partitions do not have Maxvalue Partition ● Naming & Tablespace Constraints Interval Partitioning MAXVAULE Range Based Partitioning Jan19 Feb19 Jan19 New Record with order_date of 02-Mar-2019 Interval Based Partitioning
  • 11.
    11 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 11 ● Extension of Range Based Partitions (Introduced in 11g) ● Auto creates partitions when data exceeds all the existing ranges ● Subtle Differences ● Interval Based partitions do not have Maxvalue Partition ● Naming & Tablespace Constraints ● Range Based partitions do not have any Lower Bounds Interval Partitioning MAXVAULE Range Based Partitioning Interval Based Partitioning Jan19 Feb19 Jan19 SYS_NAME New Record with order_date of 02-Mar-2019
  • 12.
    12 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 12 Partition Bounds MAXVAULE Range Based Partitioning Values less than Feb 2019 Values less than Apr 2019 Values belonging to Jan 2019 Values belonging to Apr 2019 Interval Based Partitioning New Record with order_date of 02-Mar-2019
  • 13.
    13 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 13 ● Extension of Range Based Partitions (Introduced in 11g) ● Auto creates partitions when data exceeds all the existing ranges ● Subtle Differences ● Interval Based partitions do not have Maxvalue Partition ● Naming & Tablespace Constraints ● Range Based partitions do not have any Lower Bounds Be wary of jobs that do not query data dictionary views for partition names, when moving from range to interval partitioning. Interval Partitioning
  • 14.
    14 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 14 ● Automatic List Partitions ● Extension of List Based Partitions ( Introduced in 12.2 ) ● Auto creates partitions when no corresponding partition key value found. ● The decision is to choose between o List + Default o Auto List ● Multi-Column List Partitioning ( Introduced in 12.2 ) Only one Default bucket List Partitioning Enhancements Not Suitable when too many distinct values expected.
  • 15.
    15 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 15 ● Automatic List Partitions ● Extension of List Based Partitions ( Introduced in 12.2 ) ● Auto creates partitions when no corresponding partition key value found. ● Not Suitable when too many distinct values expected. ● The decision is to choose between o List + Default o Auto List ● Multi-Column List Partitioning ( Introduced in 12.2 ) Only one Default bucket List Partitioning Enhancements D E F A U L T ASIA ASIA EMEA INDIA NEPAL U.K.
  • 16.
    16 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 16 ● Referential Partitioning ● Child table inherits the partitioning strategy of parent table through Referential Constraints ● Reduces Data Maintenance Overhead ■ A split operation on Parent will split the dependent child partitions. ■ Dropping a partition in parent table, will drop the dependent child partition – No need to disable the FK-PK ■ Applies to Truncate as well. (TRUNCATE CASCADE) ● Virtual Column Based Partitioning ● Partition based on the virtual columns of a table. Other Extensions
  • 17.
    17 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 17 ● What do you want to achieve ? ● Performance  Partition Pruning  Partition-wise join  DML Speed up  Data Load ● Data Management Benefits ● Do you know your data ? ● Type of Data - Records, transactions… ● Data Access Patterns ● Data Lifecycle Requirements. Choosing a Partitioning- Strategy
  • 18.
    18 © The PythianGroup Inc., 2018 Partition Manageability + New Features
  • 19.
    19 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 19 ● Partition a Stand-Alone Table 1. Create a separate Partitioned Table + Transfer Data 2. Exchange Partitioning ■ Create Separate Tables with Filtered Data ■ Transfer the Filtered Data into a corresponding partition 3. DBMS_REDEFINITION Online Conversion of a Non-Partitioned to a Partitioned Table CREATE TABLE ORDERS_2018 where Order_DATE < 2018 ; Alter table ORDERS_PART Exchange partition ORDERS_PART2018 with table orders_2018;
  • 20.
    20 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 20 Online Conversion of a Non-Partitioned to a Partitioned Table
  • 21.
    21 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 21 ● Global partitioned ● Non-prefixed indexes ● Prefixed indexes ● Bitmap indexes Online Conversion of a Non-Partitioned to a Partitioned Table No Effect Global Non-partitioned Indexes Local Partitioned Indexes Local Partitioned Indexes Index that contains the leading column of the partitioning key Partitioning_key => order_date Index => (order_Date, region)
  • 22.
    22 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 22 Online Split & Move Operations
  • 23.
    23 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 23 Online Split & Move Operations • Indexes always remain ONLINE • Concurrent Query & DML Operations allowed – not preferred => Impacts compression ratio
  • 24.
    24 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 24 Create Table For Exchange TAB ORA-14097: column type or size mismatch … CREATE TABLE TAB FOR EXCHANGE WITH TABLE PART_TAB; PART_TAB
  • 25.
    25 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 25 Data Management Jan19 Feb19 Mar19 Table Size 300 GB Table Details Monthly Range-based Partitions Requirements • Compress All Data older than 3 months (~25 GB/Month) • Move data older than a year from SSD To HDD
  • 26.
    26 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 26 Data Management Jan19 Feb19 Mar19 Requirements • Compress All Data older than 3 months (~25 GB/Month) • Move data older than a year from SSD To HDD • ALTER TABLE ORDER_HISTORY MOVE PARTITION OH_MAY2019 COMPRESS ; • ALTER TABLE ORDER_HISTORY MODIFY PARTITION OH_MAY2019 REBUILD UNUSABLE LOCAL INDEXES; Compression Steps • ALTER TABLE ORDER_HISTORY MOVE PARTITION OH_JUN2018 TABLESPACE ORDER_HISTORY_SLOW ; • ALTER TABLE ORDER_HISTORY MODIFY PARTITION OH_JUN2018 REBUILD UNUSABLE LOCAL INDEXES TABLESPACE ORDER_HISTORY_SLOW ; Data Archiving Steps
  • 27.
    27 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 27 Data Management Manual Operation Involved Repetitive - Every Month Multiple Partitioned Tables
  • 28.
    28 © The PythianGroup Inc., 2018 Automatic Data Optimization (ADO)
  • 29.
    29 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 29 Heat Maps
  • 30.
    30 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 30 Heat Maps ● Track Data Access ● Reads vs Writes ● Types of Reads – FTS vs Index Lookups ● Row, Segment, Tablespace Level ● Access the information via Data Dictionary Views ● Database-aware ● Maintenance jobs, stats gathering, backups, don’t affect heat map. ● Doesn’t track objects in SYSTEM and SYSAUX tablespaces HEAT_MAP = ON Advanced Compression License
  • 31.
    31 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 31 Heat Maps
  • 32.
    32 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 32 Heat Maps
  • 33.
    33 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 33 Heat Maps
  • 34.
    34 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 34 Heat Maps
  • 35.
    35 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 35 Heat Maps
  • 36.
    36 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 36 Automatic Data Optimization (ADO) ● Policies that Manage Data Automatically Compression Data Movement HOW ??
  • 37.
    37 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 37 Automatic Data Optimization (ADO) ● Policies that Manage Data Automatically Compression Data Movement HOW ?? Moving to tablespaces on separate storage tiers.
  • 38.
    38 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 38 Automatic Data Optimization (ADO) ● Policies that Manage Data Automatically Compression Data Movement HOW ?? Data can be compressed at Row storage level Segment level Group (multiple objects) level
  • 39.
    39 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 39 Automatic Data Optimization (ADO) ● Policies can be ● Space Based  E.g. When tablespace utilization reaches 85%, then move to another tablespace ● Time And Access method Based  E.g. Compress any row that has not been updated for 90  E.g. Move a segment that has not been updated for 180 to another tablespace TRIGGERS
  • 40.
    40 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 40 Automatic Data Optimization (ADO) – Time & Access Method Condition of the Policy
  • 41.
    41 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 41 Automatic Data Optimization (ADO) – Time & Access Method Goal of the Policy
  • 42.
    42 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 42 Automatic Data Optimization (ADO) – Time & Access Method Scope of the Policy
  • 43.
    43 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 43 Automatic Data Optimization (ADO) – Time & Access Method
  • 44.
    44 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 44 Automatic Data Optimization (ADO) – Space Goal of the Policy Scope of the Policy Condition of the Policy???
  • 45.
    45 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 45 Automatic Data Optimization (ADO) – Space Condition of the Policy??? • Global Defaults • Can be controlled dbms_ilm_admin.customize_ilm
  • 46.
    46 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 46 Automatic Data Optimization (ADO) - Execution ● Row Level ● Remaining Policies ● Supports Manual Evaluation & Execution Setup custom jobs for larger tables that evaluate the policy periodically MMON (15 mins) Maintenance Window
  • 47.
    47 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 47 Manual Execution
  • 48.
    48 © The PythianGroup Inc., 2018 Some More New Features
  • 49.
    49 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 49 ● Protect Data from accidental modifications ● Columns can be added /modified ● RO Partitions Cannot be Dropped ● Benefits of RO Partitions vs RO Tablespaces ● No Need for separate tablespaces ● Benefits smaller partitions Read Only Partitions ALTER TABLE ORDERS MODIFY PARTITION ORDERS_2018 READ ONLY;
  • 50.
    50 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 50 ● Applicable to Partition Move, Split, Merge ● Only rows matching the filter condition are retained Filtered Partition Maintenance Operations
  • 51.
    51 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 51 ● Partition External Tables ● Combines of Regular Tables + External Partitioned Tables Hybrid Partitioned Tables
  • 52.
    52 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 52 ● Change the partitioning strategy ● Avoids Manual Rebuild ● Uses the MODIFY Clause Modifying Partitioning Strategy
  • 53.
    53 © The PythianGroup Inc., 2018 © The Pythian Group Inc., 2019 53 Modifying Partitioning Strategy
  • 54.
    54 © The PythianGroup Inc., 2018 54 © The Pythian Group Inc., 2019 THANK YOU You can reach me : @aishwaryakala13

Editor's Notes

  • #49 We have added additional slides for use in external presentations. Client Snapshots, like the ones found in this section are also available in the sales portal and slides can be provided by reaching out to marketing.