Anda di halaman 1dari 26

Teradata Basics flashcards | Quizlet https://quizlet.

com/46137281/teradata-basics-flash-cards/

Like this study set? Create a free account to save it.

Terabyte Trillion bytes of data


10^12

Petabyte Quadrillion bytes of data


10^15

Data Warehouse Central, enterprise-wide database


that contains information extracted
from the operational data stores.

Active Data Warehousing Timely, integrated, logically


consistent store of detailed data
available for strategic, tactical driven
business decisions

Business requirements for Active Data Performance


Warehousing Scalability
Availablity
Data Freshness

Each row in the table is uniquely Primary Key


identified by a

A table may have 1 or more Foreign Keys

Can a primary key be null? No

Can a primary key contain duplicate No


values?

Can a FK have duplicate values? Yes

Does an FK allow null? Yes

1 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

TD DB Competitive advantages -Unlimited, Proven Scalability


-Unlimited Parallelism
-Mature Optimizer
-Models the Business
-Provides a "single version of the
business"
-Low TCO
-High Availability
-Parallel Load and Unload Utility

Which feature allows the TD DB to Parallelism


process enormous volumes of data
quickly?

The TD DB is primarily a Server

Major components of TD -Parsing Engines (PE)


-Message Passing Layer (MPL)
-Access Module Processors (AMP)
-Virtual Disks (Vdisk)

Parsing Engine Interprets SQL requests, receives


input records and passes data.

Message Passing Layer Made up of BYNET and PDE


All communication between PEs and
AMPs is done via the MPL

Access Module Processor Responsible for managing a portion of


the database. An AMP will control
some portion of each table on the
system. Do all the physical work
associated with generating an answer
set including, sorting, aggregating,
formatting and converting

Virtual Disk Disk space associated with an AMP.


Tables/data rows are stored in this
space. A Vdisk is usually assigned to
two or more disk drives in a disk
array.

TD Storage Procedure -The PE dispatches request to insert a


row
-MPL ensures that a row gets to the
appropriate AMP
-The AMP stores the row on its
associated logical disk
-An AMP manages a logical or Vdisk
which is mapped to multiple physical
disks in a disk array.

2 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

TD Retrieval Proceeedure -The PE dispatches a request to


retrieve one or more rows
-The MPL ensures the appropriate
AMPs are activated
-The AMPs locate and retrieve desired
rows in parallel access
-MPL returns the retrieved rows to PE
-The PE returns rows to requesting
client application

Linear Scalability Allows for increased workload without


decreased throughput.

The amount of time (or money)


required to do a task is directly
proportional to the size of the system

Data Dictionary Directory (DD/D) Integrated set of system tables which


store database object definitions and
accumulate information.
Updated automatically.

DBC.TablesV Information about all tables

DBC.UsersV Information about all users

DBC.AllRightsV Information about access rights

DBC.AllSpaceV Information about space utilization

SQL consists of: -Data Definition Language (DDL)


-Data Manipulation Language (DML)
-Data Control Language (DCL)

Data Definition Language (DDL) -Defines database structures


-CREATE, DROP, ALTER

Data Manipulation Language (DML) -Manipulates rows and data values


-SELECT, INSERT, UPDATE, DELETE

Data Control Language (DCL) -Grants and revokes access rights


-GRANT, REVOKE

View pre-defined filter of existing table(s)


consisting of specified columns
and/or rows from the table(s)

Single table view -a window into an underlying table


-allows users to read and update a
subset of the underlying table
-has no data of its own

3 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Multi-table view allows users to access data from


multiple tables as if it were in a single
table.
Used for reading only, not updating.
Also called a join view

Macro -A predefined set of SQLstatements


which is logically stored in a database

HELP command Displays information about database


objects and session characteristics

SHOW command Displays the DDL associated with


database objects, displays how an
object was created.

EXPLAIN facility Generates an English translation of


the Parser's plan.

What language is used to access a SQL


teradata table?

Node The basic building block of the


hardware for a TD system. Contains
the primary processor logic (CPUs),
memory and I/O functionality.

SMP Single node system

MPP Multi-node system


Massively Parallel Processing

BYNET -Software and hardware interconnect


that provides high performance
networking capabilities to TD MPP
systems

BYNET communication techniques Point-to-point


Multi-cast
Broadcast

If a client is a mainframe system it is channel


_____ attached

If a client is PC or UNIX based it is network


_____ attached

Call Level Interface (CLI) A library of routines that resides on


the client side.

3 Major software components in Client Application


channel-attached systems CLI
TDP

4 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Client application -Your own application(s)


-Teradata Utilities (BTEQ, etc)

CLI Consists of system calls which create


sessions, allocate request and
response buffers, create and de-block
"parcels" of information, and fetch
response information to the
requesting client.

Teradata Director Program (TDP) Manages the session traffic between


the CLI and the Database.

Host channel adapter Mainframe hardware component that


allows the mainframe to connect to a
FICON, ESCON, or Bus/Tag channel.

4 Major software components: -Client Application


Network-attached systems -Teradata CLI/ODBC
-MTDP
-MOSI

Teradata ODBC or JDBC Use open standards-based ODBC or


JDBC interfaces to provide client
applications access to TD across
LAN-based environments

Micro Teradata Director Program Must be linked to any application that


(MTDP) will be network-attached to the TD DB
Performs many of the functions of the
channel based TDP including session
management.

Micro Operating System Interface Library of routines providing


operating system independence for
clients accessing the TD DB.

Gateway Software Session balancing for LAN


connections

Parsing Engines are made up of which -Session Control


software components? -the Parser
-the Optimizer
-the Dispatcher

Major functions performed by session logon and logoff


control

The Parser... Interprets the statement, checks it for


proper SQL syntax and evaluates it
semantically.

5 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

The Optimizer... develops the least expensive plan to


return the requested response set.
Fasted alternative is converted to
executable steps, to be performed by
the AMPs, which are passed to the
dispatcher

The Dispatcher... Controls the sequence in which the


steps are executed and passes the
steps onto the MPL.
Composed of execution control and
response control.

Execution control receives the step definitions from the


parser, transmits the step definitions
to appropriate AMP(s) for processing,
receives status reports from the AMPs
and passes results on to response
control once the AMPs have
completed processing.

Done by dispatcher

Response control Returns results to the users.

Input conversion is done at the ____ Parsing Engines

Output conversion is done at the AMPs


______

Teradata stores in ASCII

PE can manage up to _____ individual 120


sessions

All communication between PEs and MPL


AMPs is done via the

Broadcast Message is routed to all AMPs and PEs


on the system

Multi-Cast Message is routed to a group of AMPs

Point-to-point Message is routed to one specific AMP


or PE on the system

6 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Database Manager -resides on each AMP


-receives steps from the dispatcher
and processes the steps
has ability to lock databases and
tables, etc.
-returns responses to the dispatcher.
-provides a bridge between the
logical organization and physical
organization of the data on disks.
-performs a space management
function that controls the use and
allocation of space.

Output data conversion -Performed by AMPs


-Check the session and change the
internal, 8-bit ASCII used by Teradata
to the format of the requester.

Each session can handle ____ requests Multiple

Each amp can perform up to ____ 80


tasks in parallel

What are 2 software elements that CLI and TDP


accompany an application on all client
side environments?

Perm space Max amount of space available for


tables, stored procedures, and UDFs.
Represents a maximum limit.

Spool space Max amount of work space available


for requests

Temp space Max amount of temporary table space

Trigger One or more SQLQ statements that


are associated with a table, executed
when the trigger is "fired".

Which objects don't need perm space? Views, macros and triggers

Teradata User database with an assigned password

A user is a(n) _____ repository, a active/ passive


database is a(n) _____ repository

True or false, databases and users true


may be empty

Perm space is a ______ limit Per/AMP

Fallback protection uses ____ x the 2


perm space of no fallback

7 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Whenever the per AMP limit is database full


exceeded on any AMP, a ___________
message is granted

Spool space has a _________ limit per amp

Temp space has a __________ limit per amp

A database will always have tables false


(t/f)

a user will always have a password true


(t/f)

a user creating a subordinate user f


must give up some of his/her perm
space (t/f)

creating tables requires the definition f


of at least 1 column and a primary
index (t/f)

the sum of all user and database perm t


space will equal the total space on the
system (t/f)

the sum of all user and database f


spool space will equal the total space
on the system (t/f)

deleting a macro from a db reclaims f


perm space for the database (t/f)

The uniformity of distribution of the Primary index


rows of a table depends on the choice
of the ________

The actual distribution is determined Hash value


by the ______________ of the primary
index

When is a primary index defined? At table creation

How many columns may be combined up to 64


to create a primary index?

A NoPI choice will result in Random generator code


distribution of data between AMPs
based on ...

How can you change the choice of PI? drop and recreate the table

Accessing the row by its primary one


index value is always a _____ amp
operation

8 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

What is the most efficient way to By its primary index value


access a row?

A UPI access may return at most _____ one


row(s)

Row hash is ______ bits for a table 32


with a PI

The high order bits of a row hash are Bucket number


a __________________

Hash bucket numbers are used to a hash map entry


identify _________________

When a system grows and new AMPs hash map


are added, this requires a change to
the _________________

Hash map -Uniquely configured for each system


-An array of entries (buckets) which
associates bucket numbers with
specific AMPs

What references the logical location Row Hash + Table ID


on an AMP?

How many bytes long is an entry in a 2 bytes long


hash map?

Hash synonym Different primary index values input


create the same row hash

NUPI duplicate Same PI value but different row will


have same row hash in a NUPI table

Row ID Row hash + uniqueness value

Only the Row hash portion of Row ID primary index, secondary index
is used in _______________ operations.
Entire Row ID is used for
______________ support

Rows are stored in a data block, row id


________________________ in sequence
by _____________________.

How to locate a row on an AMP using Start with Table ID and row hash,
a PI apply to master index to find cylinder
number. Then, taking cylinder
number, table id and row hash, apply
to cylinder index to find data block
address. Take row hash and PI value,
apply to data block to find data row.

9 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

UPI and NUPI equality value accesses true


are always a one-AMP operation (t/f)

UPI and NUPI indexes allow NULL in a true


primary index column (t/f)

UPI, NUPI and NoPI tables allow false


duplicate rows in the table

A UPI can be used as a primary key true


implementation

The output of the hashing algorithm row hash


is called the

To determine the target AMP, the MPL hash bucket number


must lookup an entry in the hash map
based on the _____________

A row ID consists of a row hash plus a uniqueness


_____________ value

A uniqueness value is required to hash synonyms and NUPI duplicates


produce a unique Row ID because of
___________ and _________

Once the target AMP has been master index


determined for a PI search, the
________________________ for that AMP
is accessed to determine the cylinder
that may hold the row.

The cylinder index points us to the Data block


address and length of the
__________________-

3 ways to access a table Primary index access


Secondary index access
Full table scan

A table can have from ____ to ___ 0 to 32


secondary indexes

Secondary index -alternate path to the data


-improve performance
-allows user to locate rows
-no influence in distribution

Secondary indexes add ____ to the overhead/ both in terms of disk space
table... and maintenance

When is a secondary index defined? at table creation or following table


creation

10 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

A secondary index can be up to 64


___________ columns

Accessing a row via a USI is a __ AMP 2


operation

Accessing row(s) via a NUSI is a ___ all


AMP operation

Creating a secondary index causes internal sub-table (deleted when


a(n) ________ to be built secondary index is dropped)

2 purposes of a USI -Speed up access to a row


-Enforce uniqueness on a column or
set of columns

Can secondary indexes be no


partitioned?

Full table scans -Every row of the table is read


-All AMPs scan their portion of the
table in parallel

Teradata executes on a ____ bit 64


operating system

AMPs and PEs are implemented as Virtual processors


_______

Shared Nothing architecture each AMP has its own memory,


manages its own disk space, and
executes independently of other AMPs

Which operating system does SUSE Linux - SLES 10 or SLES 11


Teradata 15 run under?

AMPs are associated with _________ virtuall disks/ logical units (LUNs)
which are associated with __________
within a disk array.

A multi-node system running the TD MPP (Massive parallel processing)


DB is referred to as a _____ system

TPA node -Trusted parallel application node


-executes teradata database software

HSN -Hot standby node


-Is a spare node in the clique (not
running teradata) used in event of a
node failure

Non-TPA node -application node that does not


execute TD DB software

Clique group of nodes connected to and


sharing the same storage

11 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

2+1 clique 2 nodes + hot standby node

point-to-point bynet communication -one to one


-one vproc communicates with one
vproc

multi-cast bynet communication -one to many


-one vproc communicates to a subset
of vprocs

broadcast bynet communication -one to all


-one vproc communicates to all
vprocs

the software component that allows PDE


the TD DB to execute in different
operating system environments is the
_____

A physical message passing bynet


interconnect is called the _____

a clique provides protection from a node


_____ failure

If a node fails, all vprocs will migrate vproc migration


to the remaining nodes in the clique.
this feature is referred to as

The _____ or ____ provides a single SWS/ AWS


point of operational control for
Teradata MPP systems.

A ____ node is part of a system TPA


configuration, is connected to the
BYNET and executes the teradata
database software

A _____ node is part of a system NOTPA


configuration, connects to the BYNET
and is used to execute application
software other than TD DB software

A ____ node is part of a system HSN


configuration, connects to the BYNET,
and is used as a spare node in the
event of a node failure

Fallback provides data access with a AMP


"down" ____

Locks provide data ______ integrity

Transient journal provides automatic aborted transactioins


rollback of ______

12 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Permanent journal optional before and after-image


journaling

4 types of locks -exclusive


-write
-read
-access

Down-AMP recovery journal started automatically when the system


has a failed or down AMP. Its purpose
is to log any changes to rows which
reside on the down AMP

Most commonly used RAID in 1


Teradata?

RAID 1 -data mirroring protection


-requires each primary data disk to
have a companion disk or mirror. The
contents of these two are identical
-Transparent mirroring
-Storage costs are high

A large clique is usually a set of ___ 8 (large cliques no longer supported)


Teradata nodes

Fallback cluster defined set of 2 AMPs across which


fallback is implemented

Loss of one AMP in the cluster... permits continued table access

Loss of two AMPs in the cluster.... causes the RDBMS to halt

AMP clusters are defined as ___ AMPs 2

RAID 1 mirroring or RAID 5 data parity disk drive failure


protection provides protection in the
event of _________

Failures fallback provides protection -multiple drive failures in the same


against drive group
-an array is not available
-an amp is not available

If one disk drive fails in the first drive No, mirrored disk handles this
group is fallback needed?

If one disk drive fails in the first group No, mirrored disks handle this.
and one fails in the third group is
fallback needed?

If 2 disk drives fail on the first drive Yes.


group is fallback needed?

Exclusive lock prevents any other type of concurrent access

13 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Write lock prevents other reads, writes and exclusives

Access lock prevents exclusive only

Read lock prevents writes and exclusives

Select automatically applies a ____ read


lock

Update automatically applies a ____ write


lock

Create table automatically applies a exclusive


____ lock

Transient Journal -exists to permit the successful


rollback of a failed transaction
-Maintains a copy of all before images
of all rows affected by the transaction.
If a transaction fails, the before
images are reapplied to the affected
tables, the images are deleted from
the journal and a rollback operation is
completed.

Where are before images (UNDO rows) within WAL


maintained?

Is the recovery journal automatic? yes

Where is the recovery journal by other AMPs in a cluster


maintained?

ARC utility -allows for the archiving and


restoring of database objects which
may have been damaged or lost
-Archives and restores data from/to
Teradata DB
-Restores or copies data from archive
media
-Permits data recovery to a specified
checkpoint using Permanent Journals

BAR -Backup and recovery architecture


-provides solutions from Teradata
Partners
-Provides Windows front end for ARC
-Easy creation of scripts for
archive//recovery
-Ex: NetVault, NetBackup

What provides for TXN rollback in Transient Journal


case of failure?

14 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Teradata backup and recovery NetBackup/Tivoli


applications...

Protects all rows of a table Table locks

Logs changed rows for down AMP Down AMP recovery journal

Provides for recovery to a point in Permanent Journal


time

Applies to all tables and views within Database locks

Multi-platform archive utility ARC

Lowest level of protection granularity row hash locks

Protects tables from AMP failure Fallback

Protects database from a physical Disk Array


drive failure

Group of AMPs used by fallback cluster

4 stages of Database Development -Requirement analysis


-Logical modeling
-Activity modeling
-Physical modeling

Requirement analysis eliciting the initial set of information


and processing requirements from
users

Logical modeling determines the contents of a database


independent of a particular physical
implementation's exigencies
-Conceptual modeling and view
integration

Conceptual modeling transforms the user requirements into


a number of individual user views
normally expressed as entity-
relationship diagrams

View integration Combines these individual user views


into a single global schema expressed
as key tables. The logical model is
implemented by taking the conceptual
moodel as input and transforming it
into the data model supporting the
target RDMS. The result is the
relational data model.

15 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Activity modeling Determines the volume, usage,


frequency, and integrity analysis of a
database. This process also consists
of placing any constraints on domains
and entities in addition to addressing
any legal and ethical issues including
referential integrity.

Physical modeling Transforms the logical model into a


definition of the physical model
suitable for a specific software and
hardware configuration. In relational
terms, this is usually some schema
expressed in a dialect of the data
definition language of SQL.

Business Information Model (BIM) -shows major entities and their


relationships

Logical Data Model (LDM) -3rd normal form


-BIM plus all tables, minor entitiies,
PK-FK relationships, constraints and
attributes

Extended Logical Data Model (ELDM) LDM plus demographics and


frequencies

Physical Data Model (PDM) ELDM plus index selections and any
denormalizations

Entity-Relationship diagram depicts entities and relationships

Relational acronym: PK Primary key

Relational acronym: FK Foreign key

Relational acronym: SA System assigned

Relational acronym: UA User assigned

Relational acronym: NN No NULLs

Relational acronym: ND No duplicates

Relational acronym: NC No changes

Derived attributes -An attribute whose value can be


calculated or otherwise derived from
other attributes
-Creates redundant data

What are the three types of 1:1, 1:M, M:M


relationships?

Decomposable data data than can be broken down into


finer, smaller units

16 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Atomic data data that is already at its finest level

When comparing character data to character to numeric


numeric, Teradata will always convert
_______ to ______ then do the
comparison

First Normal Form (1NF) Attributes must not repeat within a


table. No repeating groups.

Second Normal Form (2NF) -An attribute must relate to the entire
primary key, not just a portion.
-Tables with a single column PK
(entities) are always in 2NF

Third Normal Form (3NF) -Attributes must relate to the PK and


not to each other
-Cover up the PK and the remaining
attributes must not describe each
other

Normalization The process of placing attributes in


the correct entities

Benefits of normalization -Minimize redundancy


-Provide optimum flexibility
-Eliminate update anomalies

Costs of normalizing to 1NF include -You use more disk space


-You have to do more joins

Costs of normalizing to 2NF when You have to do more joins


already in 1NF include

Costs of normalizing to 3NF when You have to do more joins


already in 2NF include

Which data model would include the Physical data model


definition of a partitioned primary
index?

Which two factors should be Volatility and performance


considered when deciding to
denormalize a table?

What is a benefit of implementing avoid data conversion


data types at the domain level?

What are the three primary index -Access Demographics


choice criteria? -Distribution Demographics
-Volatility

Access column a column that would appear in a


where clause in a SQL statement

17 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

DBSControl General Flag Primary -Default setting


Index Default: "D" -First column selected as a NUPI

DBSControl General Flag Primary -First column selected as a NUPI


Index Default: "P"

DBSControl General Flag Primary The table will be created without a


Index Default: "N" primary index (NoPI table)

Advantage of multi column primary uniqueness


index: more columns = more
_________________

Disadvantage of multi column primary Usability


index: more columns = less
_________________

How can you maximize one-AMP By basing your primary index on the
operations by choosing primary column(s) most often used for access
indexes?

The number of distinct values for a PI AMPs


must be greater than the number of
______ in the system.

____ tables do not allow duplicate Set


rows

_____ tables do allow duplicate rows Multiset

To avoid NUPI duplicate row checks, Multiset


create the table as a _______ table

When a row is inserted into a Set table NUPI duplicate row check
with a NUPI, the system must perform
a ___________

PI candidate: Distinct values- the more


_____ the better

PI candidate: Max row per value- the fewer


_____ the better

PI candidate: max rows null- the fewer


______ the better

PI candidate: typical rows per value- fewer


the ______ the better

DBC.TableSizeV Provides AMP Vproc disk space usage


at the table level.

Identity column -causes the system to generate a


table-level unique number for the
column for every inserted row,
whether for single or bulk inserts

18 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

____ assigns identity column values PE


for single inserts and TPT Stream
(Tpump)

_____ assigns identity column values AMP


for TPT Load (FastLoad), TPT Update
(MultiLoad), and SQL Insert/Select
operations

How/ where are identity columns At column level in a CREATE TABLE


implemented? statement

Identity column: CYCLE restarts numbering after the


maximum/minimum number is
generated

How many identity columns can a 1


table have?

Can you add an identity column to an no


existing table with an ALTER TABLE
statement?

What is the maximum numeric data DECIMAL (18,0) and NUMERICA(18,0)


type range for identity columns?

NoPI tables are a Teradata _____ 13.0


feature

Where are rows added when inserted At the end of the table
into a NoPI table?

Loads using TPT Load (FastLoad) or NoPI staging table


TPT Stream (TPump) array insert into
a _______ are faster

What clause do you specify in the NO PRIMARY INDEX


CREATE TABLE statement to create a
NoPI table?

TableKind for a NoPI table O

A NoPI table is automatically created Multiset


as a _____ table

For SQL INSERTS into a NoPI table, the PE, random generator code
____ uses ______________

For TPT load or FastLoad inserts, how Blocks of data are distributed between
are rows distributed for a NoPI table? the AMPs in a round robin fashion

For a NoPI table, HBN is how many 20


bits?

For a NoPI table, uniqueness value is 44


how many bits?

19 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Row ID for NoPI is made up of a ___ 20 bit hash bucket number, and a 44
bit _______, and a _____ bit ________ bit uniqueness value

Row ID for a PI table is made up of a 32 bit row hash, and a 32 bit


_____ bit ____, and a _____ bit _______ uniqueness value

Is Archive/Restore supported for NoPI Yes


tables?

What is a problem with restoring a Data in a NoPI table can be skewed


NoPI table? after a restore or copy.

Is FALLBACK available with NoPI Yes


tables?

Is permanent journaling allowed with no


NoPI tables?

Are identity columns allowed with no


NoPI tables?

Which trade-off must be balanced to Access and distribution


make the best choice for a primary
index?

When volatility is considered as one of How often the data values will change
the primary index choice criteria, what
is analyzed?

To optimize the use of disk space, the is unique or nearly unique


designer should choose a primary
index that...

For NoPI tables, what are 2 ways in 1) At the AMP level with TPT load
which the Random Generator is 2) At the PE level for ad hoc SQL
executed? requests

What is row partitioning (PPI)? -an indexing mechanism in TD for


use in physical database design
-Data rows are grouped into
partitions at the AMP level -
partitioning is simply an ordering of
the rows within a table on an AMP

What advantages does partitioning -Increases available options to


provide? improve performance of certain types
of queries (specifically range-
constrained queries)
-Only the rows of the qualified
partitions in a query need to be
accessed - avoid full table scans

20 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Disadvantages of partitioning -Partitioned rows are 2 to 8 bytes


longer
-PI access may be degraded if the
partitioning column isn't part of the PI
-Joins to non-partitioned tables with
the same PI may be degraded
-The PI can't be defined as unique
when the partitioning column is not
part of the PI

A table has a max of ____ partitions 9.223 quintillion

CASE_N is patterned after SQL _____ CASE


expression

If the partitioning column(s) are not unique


part of the PI, the PI cannot be
_________

To maintain uniqueness on the PI of a USI


partitioned table, you can create a
____ on the PI

Before Teradata ___, the max 14.0/ 65,535


partitions was ________

For empty tables with a PPI, the ALTER -remove partitioning for a artitioned
TABLE statement can be used to: table
-establish partition for a table
-change the columns that comprise
the PI
-change a unique primary index to
non-unique
-change a non-unique primary index
to unique

If the partitions have data, which WITH DELETE


syntax statements are required? or WITH INSERT

What happens to the data in the the data in the dropped partitions is
dropped partitions when you drop also deleted form the table
partition ranges in a PPI table?

If NO RANGE is defined, if the Moved to NO RANGE partition


dropped partitions have data, what
happens to the data in said partitions?

If WITH DELETE or WITH INSERT syntax it is effectively ignored


is included, assuming NO RANGE is
defined...

21 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

If new partitions are added if NO NO RANGE partition is checked and


RANGE is defined, what happens? rows that apply to the newly aded
partitions are moved from NO RANGE
into the new partitions

If not on first level of partitioning, #L2,3,4,etc...


how do we reference the level??

The feature that allows tables and CPPI


non-compressed join indexes to
include partitioning on a character
column is referred to as....

Prior to TD 13.10, partitioning numeric or date


expressions were limited to what
types of data?

In a PPI table, every row is uniquely Partition number + row hash +


identified by its.... unique value

The row key consists of partition number + row hash

In a NPPI table, the partition number 0


defaults to...

T/F, for a PPI table, the partition false


number and the row hash are both
used by the message passing layer to
determine which amps should receive
the request

A USI may be used to maintain uniqueness


__________ on a column

Can SI subtables be accessed by users no


directly?

Which two ways can secondary -when a table is created (CREATE


indexes be defined? TABLE)
-for an existing table (CREATE INDEX

PI access is a ___ AMP operation one

USI access is a ___ AMP operation two

NUSI access is a ___ AMP operation all

USI subtable rows are distributed by Row Hash


the ____

Max size of a single nusi row 64KB

If the base table is Fallback, will the Yes


secondary index subtable be Fallback?

Does the optimizer automatically use no, it depends.


a NUSI

22 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Without COLLECTed STATISTICS, FTS


optimizer often chooses to do a (FTS
or NUSI)

Hash-Ordered NUSIs can be used value access


for... join access
covering index

Value-Ordered NUSIs can be used for value access


range access
covering index

NUSI Bit Mapping an optimizer technique to determine


common Row IDs between multiple
NUSI values

The Optimizer is more likely to COLLECT STATISTICS


consider bit mapping if you_____

NUSI Bit Mapping requires at lease __ 2


NUSI equality conditions

A value-ordered NUSI is limited to a -single numeric (4-byte)


_______ or ______ column -DATE

How are value ordered NUSI's index by data value


subtable rowws sorted?

VONUSIs provide major advantages in range


performance of ____ queries

The Optimizer generally chooses a -statistics have not been collected on


FTS over a NUSI when: the NUSI
-The index is too weakly selective
-small tables

_____ is a process that determines NUSI bit mapping


common Row IDs between NUSI
values by a process of intersection

Covering Index If aggregation is performed on a NUSI


column, the Optimizer accesses the
NUSI subtable and returns the result
without accessing the base table

Do USIs support nested joins? yes

Do USIs support row hash locking? yes

A UPI is also a ____ candidate and a USI/NUSI


NUPI is a ____ candidate

Value access How often a column appears with an


equality value

23 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Range access How often a column is used to access


a range of date values

To be considered as a NUSI candidate, 20


a column shoudl have at least ___
distinct values

Which 2 types of additional overhead -additional I/O maintenance


do SIs have? -Secondary index subtables require
additional disk space

Transient Journal provides _____ Transaction


integrity

Where are Transient Journal images WAL log


maintained?

Transient Journal provides for _____ in automatic rollback


the event of TXN failure

The WAL log includes: -Before images or undo records used


for transaction rollback
-After images or redo records for
updating disk blocks

Multi-Table Join Index Pre-join multiple tables


Stores and maintains results from
joining two or more tables

Single Table Join Index Distribute the rows of a single table


on the hash of a foreign key value

AJI -aggregate one or more columns of a


single table or multiple tables into a
summary table

Sparse join index Join index with a WHERE clause

Global join index You can include the Row ID of the


tables within the join index to allow
an amp to join back to the data row
for columns not referenced in the join
index

Are join indexes updated Yes


automatically as base tables are
updated?

What type of space do JIs occupy? PERM space

Can BLOB/CLOB data types be defined No


within a join index?

What type of join can you not specify Full outer


in a multi table join index?

24 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Can you create USIs on JI columns? USIs- No


NUSIs? NUSIs- Yes

Hash Indexes are created for a single True


table only (t/f)

Can a trigger and a hash index exist No


on a table?

Can USIs/NUSIs be built on hash No


indexes?

Hash indexes are always defined on a t


single table (t/f)

Are aggregation operators permitted No


with Hash Indexes?

Teradata allows for sessions to BTET (Teradata) mode


operate in either... ANSI mode

BTET uses ____ mode Teradata

ANSI uses _____ mode ANSI

BT begins an explicit transaction

ET commits the currently active


transaction

Transactions are by definition implicit Teradata


in ____ mode

Transactions are by definition explicit ANSI


in ____ mode

(t/f) to change session mode, you true


must logoff first

What is ANSI standard: DATE or CURRENT_DATE


CURRENT_DATE

Types of temporary tables Derived tables


Volatile Tables
Global Temp Tables

Derived tables are materialized in ____ SPOOL


space

Volatile tables use ____ space SPOOL

Global Temp Tables use ____ space TEMP

Do derived tables have data dictionary No


involvement?

Do Volatile tables have data dictionary No


involvement?

25 of 26 10-12-2015 10:57
Teradata Basics flashcards | Quizlet https://quizlet.com/46137281/teradata-basics-flash-cards/

Do Global temp tables have data Yes


dictionary involvement?

Glocal temp tables are discarded at session end


when?

Volatile tables are discarded when? at session end

which commands will not work in BT,ET


ANSI mode?

Which BTEQ setting controls td vs .SET SESSION TRANSACTION


ANSI?

26 of 26 10-12-2015 10:57

Anda mungkin juga menyukai