Data Recovery And Protection in Teradata

 

Locks:- Prevent simultaneous access to the objects

RAID:- Prevent from disk failure

Fall back:- Prevent from Amp failure

Journal:- Prevent from Image failure

CLIQUE:- Prevent from Node failure

 

ARC[Archive and Recovery]

A Means of Archiving data to tape and disk and restoring data to teradata database backup application software includes the following

Teradata archive and recovery, Net Valet, Net Backup, Tivoli Storage Manager,  Teradata Extension.

Note:- the above all feature are provides teradata, to makes teradata as highly

LOCK Held	Table Name	“A” having Lock Access	Read	Write	Exclusive
Access	√	√	√	√	X
Read	√	√	√	X	X
Write	√	√	X	X	X
Exclusive	√	X	X	X	X

 

Inclined to build a profession as Teradata Developer? Then here is the blog post on, explore Teradata Training

Access

DATA Inconsistent, select

Read

DATA consistent, select

Write

Insert, Update, Delete

Exclusive

DDL Statement

HUT LOCK

•  This is the lock applied while we are doing archive and recovery.

Syntax- Locking row for

<Lock type><Sql Queue >

Example

Locking row for access SEL* From party Locking row for write SEL*from party

Syntax Table level

Locking table<Table name> for

<Lock type>

Example Locking table party for access

Release Lock

Syn Release Lock database name Table name:

Advantages of Access Lock

a) Permit queue access to a table multiuser Environment

b) Having minimal blocking efforts on other queries

c) Very use full for aggregating a huge number of rows.

Access Lock applied Immediately even though there are many locking it’s that query

Note- Refer to the complete description And practical page NO:50 to 52 new material

RAID-[REDUNDENT ARRAY OF INDEPENDENT DISK DRIVE] Refer to the page No: 53 To 54 for the description

RAIDO,RAID1,RAID2,------RAID5,RAIDS,RAIDE etc.,

Teradata supports

RAID1, RAID5, RAIDS

RAID1:[MIRRORING TECHNIQUE] Real-time use

Advantage:-quick Availability, primary disk failure, recovers a/s easier

Drawback

1. 50% Memory wastage
2. Operations double[Insert, update, Delete] 

RAID5:-[PARTY CHECKING TECHNIQUE]

P= B1+B2+B3

B1=P- B2-B3

Advantage 25% only memory addition[less memory]

Drawback: Recovery slow

Fall back:- Refer to the page no:54

Usage: At the time of AMP failure, it permits to access rows in others AMP’S

Party:            PID  PNM

1                               X

2                              Y

3                              Z

4                             M

Advantage: High availability or fast data[At time of failure, immediate redirection to fall back AMP, so that data highly accessible]

Drawback

1. 50% Space waste
2. Operations double

Note: In real-time we use raid1 and fall back as part of data protection and recovery general we back cluster environment in real-time

Fall back Cluster(3,4,8,16)

Usage: Cluster is nothing but a group of Amp’s, then it takes a single place, where fall back maintain with in the same group

It is an advantage because we will skip scanning other Amp’s And other groups

Note: If 2 Amp’s went down RDBMS halted, Tera data recommend 4 Amp’s for the cluster for detail description refer to page No:55 To get average performance

Journal: Journal is a kind of record, which close some kind of activity

No. of Amp(N)	Workload1+1/N-1	Remains should available
16	1.06	15
8	1012	7
4	1.3	3
2	2	1

Journal Types

• Recovery journal
• Transient Journal
• Permanent journal

Single image  one copy of data

Dual image  two copies of the data

Before image before the change

Afterimage  after change taking the image

Recovery Journal

1. Automatic[of fall back is there]

Usage Once the down Amp came online[repaired] to recovery successfully, recovery journal log are useful

Teradata Interview Questions

2.Transcient journal

 [it helps is to do successful]

• Transaction nothing but a logical collection of the statement, which can be either failure or successful
• Here transaction journal helps to do a successful roll back in case of transaction failed.

Successful Transaction                                    Failed Transaction

BEGIN Transaction;//;Transient             _________________  journal activated(T.J)

1. S) INSERT PARTY;//Back up image ____________ B.I (B.I)taken

1. S) UPDATE PARTY; ________(party)
2. S) Delete Party; ________(failed)

END TRANSACTION;

1.Drops B. I                 B. I dropped               1.Roll back happens

2.Discards T.J            T.J discarded             2.B.I replied

Note: Transient journal uses before image for its operations

 3.Permanent journal

• It helps us to do fall(or)partial selective recovery
• It uses both before and after image[single and clual]

Ex: Till yesterday the data was corrupted

Till yesterday  B.I

Mon                             à A.I

|

|

|

Sun A.I

Full recovery= B.I+[Mon-Sun] A.I

Selection recovery = B.I+[Mon-This] A.I

Note:- Refer to the material description of example the material

CheckSum

1. It used to prevent disk input, output Errors
2. A checksum is completed value, from in algorithm generated by an algorithm on the data block this Algorithm applied storage and retrieval level it these are no difference, there a no disk input, output error

If x=y No, disk input, output error

If x<>y, disk input, output error

Default:- Algorithm applied on data block recording value specified DBC Control utility according to table type

Note:- Many people they using in real-time to implement Change data capturing concepts

Clique:-

• Group of nodes sharing common disk array Driver is called clique
• Each and every multi-system have at least one clique
• The nodes of clique connected to the controls of the disk array
• Each clique has the same no of nodes

Advantages:-

• It provides data availability, even the nodes failed by Migrating virtual processor[PE, AMP]
• Migrating from fail node other nodes [operation node]

Note:-

1. Multiple cliques in the same system must have the same no. of nodes
2. All these nodes communicate with each other by using by NET

HSN[HOT STAND BY NODE]

It is a node, that is a number of cliques that do not configure that execute any tera data vproc if a node in the clique fail

The Amps from the fail node move to the thigh HSN Stand  by Node in this way it decreases performance in Zero space zero percentage