Active/passive ports sorted input distict

filter Active i/o na na

Router Active i/o na na

sorter ACTIVE i/o/key/direction no yes

exp PASSIVE I/O/V NO NO

agg Active I/O/V/GROUP BY YES NO

joiner Active I/O/M YES NO

source qualifier Active I/O NO YES

Union Active I/O NO NO

rank Active I/O/V/R/ GROUP BY NO NO

Update Stratey Active I/O NO NO

Seq gen PAssive O NO NO

Connected lookup PASSIVE I/O/L NO NO

un -conn PASSIVE I/O/L/R (RETURN) NO NO

dynamic lookup PASSIVE I/O/L/ ASSOCIATED PORT NO NO

Normalizer ACTIVE I/O NO NO

Transaction Control Active I/O NO NO

cache properties Default Port

na Default condition True na

na Default condition True Default group is available

sorter cache Distinct na

NO NO NO

INDEX AND DATA SORTED IPUT NO

JOIN TYPE - NORMAL JOIN

MASTER OUTER
DETAIL OUTER
INDEX AND DATA FULL OUTER NO

SQL QUERY
USER DEFINED JOIN
SOURCE FILTER
NO.OF SORTED PORTS
DISTINCT
NO NO

NO NO NO
INDEX AND DATA TOP/BUTTOM, NO.OF RANKS RANKINDEX

DD_INSERT -0
DD_UPDATE-1
DD_DELETE-2
NO DD_REJECT-3 0

Current Value
Incremented by
End value
Reset
Cycle CURRVAL
NO start value NEXTVAL
INDEX AND DATA cache Lookup SQL override ,
static lookup source filter
dynamic enable lookup cache
persistant cache lookup policy on multiple match
shared cache dynamic lookup cache
insert else update no

STATIC CACHE WE WILL CALL LOOKUP FUNCTION NO

1.enable dynamic lookup cache

2.check insert else update NewLookupRow
3. assign the assoicated port
4. Connect associated port to the 0 - no change in the cache
next tr 1 - insert in the cache
DYNAMIC CACHE 2- update in the cache

I input and output for each field

and multiple inpiu
ts for ocuurance fileds
one GCID and one GK default
NO NORMALIZER ports

tc_continue transaction
tc_commit_before
tc_commit_after
tc_rollback_before
NO tc_rollback_After no
groups Performance

na use Filter transformation as early as possbile

instead of using muntiple filter use singled router

transformation

We can hve multiple conditions in a single router

single input , multiple
outputs and one default we can capture not satisfying records into default
group group
use operators than fucntions
instead of concat funciton use ||
na Informaticaoperations
Numerical sorter tr isare
faster than
faster oracle
than order by
string
operations
Use variable ports to reduce the repetative
NO calculations

1.USE SORTED INPUT OPTION IN AGG and send

Sorter data to agg TO AVOID THE AGG CAHCE AND
TO REDUCE TO TIME TO GROUP UP
2.GROUP BY NUMERICAL FILEDS FASTER THAN
STRING FIELDS
3.USE FILTER BEFORE AGG
4.DNT PASS UN-NECESSARY RECORDS AGG
GROUP BY PORT LIMIT THE NUMBER OF I/O

Use the less number of records tabel as master

table.
Use sorted input
use Normal join as possiblr
NO join in source qualifier whenver possible

USE SQ ALWAYS TO JOIN HOMOGENOEUS

SOURCES TABLES
USE SQ INSTEAD OF FILTER
NO USER SQ INSTEAD OF SORTER

MULTIPLE INPUT
SINGLE OUTPUT
NO

no to generate the surrogate keys , we should seq tr

 Add in index to the columns which you using
lookup condtion.
always use equality operaor first in lookup
condtion
Always use the smaller table as the lookup table
Try to do left outer join in source qialifer when the
tables in the same database.
Use persinstant lookup cache for static lookups
Use un unconneted whenver possible.
You can over ride default order by contion in
no lookup

NO YES , WHEN YOU WANTED SEND THE DIFFERENT COL

no yes , Dynamic lookup cache will change during the se

usage :
1. you can use Normaliser Tr to generate multipe
records from single record.
no 2. To read cobol sources

no no
Limitations / DISADVANTAGES
If we have mutiple filters , each filter will READ each incoming record once.
We cannot use multiple conditions in single filter TR

you cannot use group/agg functions in exp

only 4 types joins of supported

CANNOT JOIN HETEROGENIES SOURCES

UNION WILL NOT REMOVE THE DUPLICATES

CANNOT UNION IF THE TABLES ARE IN DIFF STRUCUTRE

IT WILL RETURN ONLY ONE VALUE
UN-CONNECTED LOOKUP WILL USE FOR STATIC LOOKUPS

cannot use dynamic lookup in un-conncted lookup and also for flat file lookups

you cannot reuse Normalize tr which is used to read the cobol sources

no
WHY ROUTER IS ACTIVE?
BEAUSE ROUTER IS HAVING MORE THAN ONE OUTOUT GROUP

If any transformation can change the order of records (e.g. a Sorter).

 Active/passive ports sorted input distict

Filter Active I/O NA NA

router Active I/O NA NA

sorter

exp Passive I/O/V NA NA

agg
joiner
source qualifier
Union
rank
Update Strateg
Seq gen
Connected lookup
un -conn
dynamic lookup
Normalizer
Transaction Control
cache properties Default Port

NA DEFAULT CONDIITON IS TRUE

NA NA DEFAULT CONDIITON IS TRUE

NA NA NA
groups

NO
one input , one default and multiple
output grups

NA

ORACLE
UNION WILL NOT DUPLICATES
UNION ALL WILL ALLOW DUPLICATES

WHENEVER YOU WANT TO USE SORTED AGG OR JOINER , IS IT MANDOTRY USE SORTER TR

YES
NO
 Performance

un satisfied records can caputre

we can we multiple condiotions with the single router
for each incoming , router will read only time

use Variable port to redure complexity and also it will calucate expression only time , we
can reuse the variable ports muliple times
always use nuerical funcitons than string functions
Always use operators than funcitons

INFORMATICA

UNION

SORTED AGG OR JOINER , IS IT MANDOTRY USE SORTER TR

SOURCE IS FLAT FILE ,SORTER BEFORE AGG OR JOINER

USE ORDER BY IN SOURCE QUALIFER WHEN YOUR SOURCE IS TABLE
Limitations
CAN USE ONLY ONE CODITION
un satisfied records CANNOT BE CAPUTRED
IF WE USE MULTIPLE FILTER, FOR EACH INCOMING
RECORD , IT WILL READ MULTIPLE TIMES.
Others

If any tranformation having more than oneput group or more than output group - active
1
2
3
4
5
filter
can use only one condition
we cannot capture the rows which does not meets the condition
No groups in filter
mulriple filters will read incoming records more than once.
if you use multiple filter , performance of mapping will decrease
router
can use multiple conditions
we can capture the which does not meet the conditon by using dfault gourp
single inout , multiple outs , one default group
Router will read incoming records only once.
it will improve the performance .
TABLE1 TABLE2
Inner join left outer join
sno sno sno sno sno
1 5 1 1 1
2 6 2
3 7 3
4 1 4

sno sno sno sno sno

1 1 1 1 1
2 2 2 2 2
3 4 3
5 5

sno sno Inner join left outer join

1 1 1 1 1
2 1 1 1 1
3 2 2 2 2
5 4 3
5

sno sno sno sno sno

1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 null 1 1 1
null 1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
null
left outer join right outer join
sno sno sno
1 1 1
null null 5
null null 6
null null 7

sno sno sno

1 1 1
2 2 2
null null 4
null

left outer join right outer join

1 1 1
1 1 1
2 2 2
null null 4
null

sno sno sno

1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
1 1 1
null null null
 Inner join
sno
1

sno
` 1
2

1
1
2

sno
1
1
1
1
1
1
1
1
1
1
1
1
 left outer join right outer join
sno sno sno sno sno
1 1 1 1 1
2 null null 6
3 null null 7
4 null null 5

sno sno sno sno sno

1 1 1 1 1
2 2 2 2 2
3 null null 4
5 null

1 1 1 1 1
1 2 2 1 1
2 3 null 2 2
5 null null 4
1 1

sno sno sno sno sno

1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
 1
2
3
4
5
6
7

OT_SRC_Dev3
SQ
TO JOIN HOMOGENESOUS TABLES ONLY
SQ WILL SUPPORT ALL TYPES OF JOINES WHICH SUPPORTED SOURCE DATABASE
TO JOIN N SOURCES TABLE , 1 SQ REQUIRED
WE CAN OVERRIDE SQL QUERY
WE HAVE SOURCE Filter , USER DEFINED JOIN, NO OF SORTED PORTS, DISTINCT AVAILBLE
SQ CAN SUPPORT ANY OPERATOR
USE SQ TO JOIN WHENVER POSSBILE

FLAT FILE
TABLE
JOINER
TO JOIN HETEROGENEOUS SOURCES
JOINER WILL SUPPORT ONLY 4 TYPES OF JOINS ONLY
TO JOIN N SOURCES , N-1 JOINER TR REQUIRED
THERE IS NO SQL QUERY
NONE OF PROPERTIES AVAIABLE
JOIN WILL SUPPORT ONLY EQUALITY OPERATOR
USE ONLY WHEN SOURCE ARE HETEROGENEOUS

TABLE
TABLE
 ROUTER
1 HAING ONE INPUT , ONE DEFAULT AND MULITPLE OUTGOURPS
2 ROUTER WILL SPLIT THE SOURCE DATA MULTIPLE TARGETS
3 DEFAULT GROUP
4 WE SHOULD GIVE CONDTION FOR EACH GROUP
UNION
MULIPLE INPUT GROUPS AND ONE OUTPUT GROUP
WILL COMBINE MULTIPLE SOURCES AND SEND THE DATA SINGLE TARGET
NO DEFAULT GROUP
NO CONDITIONS
 SQ
1 We can join only homogeneous source tables
2 Any join can be possible
3 no cache
4 Actice
5 I/O
6 In SQL query no alias will generate
7 SQ will not default order by
LOOKUP
can join heterogenous sources
defalt leftouter join
5 types of cache DATA CACHE
passive INDEX CACHE
I/O/L/R STATIC CACHE
but here Alias will generate DYNAMIC CACHE
Lookup will wirite default order by clause SHARED CACHE
PERSISTANT CACHE
 joiner
1 will support 4 types of joins
2 There is no option to write a sql query
3 no Source filter availble
4 2 cache
5 Active
6 I/O/M
7 JOINER WLL SUPPORT ONLY EQUALITY OPERATOR
8 JOINER IS on connected TR
lookup
only Default left outer join
We have lookup SQL override
source filter poprety availble
5 types cache
Passive
I/O/L/R
LOOKUP WILL SUPPORT OTHER THAN EQUALITY OPERATOR
lookup can be connected or un-connected TR
 union joiner
1 Multiple inut , one output group there are no groups
2 there is no condtion there is a condition

In aAll input/output Groups no need of sme, ONLY THE

Number of ports , datatype , FILEDS WHICH ARE IN JOIN
Length , precision and scale CONDTION , DATA TYPES
3 should be same SHOULD BE SAME
5 it wil not remove any duplicates it will remove duplicates
6 IT WILL ADD ROWS IT WILL ADD COLUMNS
 static lookup cache
1 static cache remains same during the session run
2 you can use euqility , >,<
3 Lookup policy on multiple match
4 flat files and relational tables
5 can be used in connected and un-connected lookups

STATIC lookup cache

1 STATIC LOOKUP WILL REMAINS SAME THORUGHT THE S
2 static cache will used for files and table lookups
3 by using static lookup we can use lookup policy on mul
4 we should use any operator > , >=, <. <=
5 static lookup can be use connected and un connected l
6 newlookup row, associated port is not avaible
Dynamic lookup cache
cache will change during the session run
equality operator
we cannot handle here
relational tables
only connected lookup

DYNMAIC lookup cache

At Run time Look cache will change , either insert or update
dynamic inly relational lookup
dynamic lookup it will disabled
dynamic lookup will use only equality operator
should use connected lookup
here it will available.
Un

Connected looukp
1 Its a part of data flow
2 will return multiple values
3 we can use static and dynamic lookup cac
4 More visible
5 For every record , lookup will execute
un connected lookup
It is separate from the data flow.
will return only one value by checking R port
we can use only static
Less visible , will call lookup tr in expression tr
only executes when lookup fucniton Is called
1
2
3
Mapplet
Mapplet should have atleast on mapplet output trasnformation
Mapplet contains mapplet input Tr and mapplet output TR
Mapplet can be re-usable across any mapping

Re-usable Tranformations
Re-usable Tranformation is a single object which re-use across the mappings

you cannot create Source Qualifier TR , Normalizer which whichis used to

read the cobol sources , ERP source qualifer as re-usable TR
mapping
Mapping should have atleast one target
Mapping contains Source , SQ, Target
A mapping cannot be re-usable

Mapplet
Mapplet is a set of tranformationS which contains a logic ebedded in it and can re-use the mapplet across any tr

Mapplet cannot include follwing Objects

1. we can use use only re-usable seq gen in the mapplet
2. we cannot use a re-usable stored procedure tranforamtion.
4.You cannot include the following objects in a mapplet: --- limitations – IMP IQ
Normalizer transformations
COBOL sources
XML Source Qualifier transformations
XML sources
Target definitions
Other mapplets
let across any tr
 Bulk
1 Data will laoded to the target table faster
2 Database logs be skipped , we cannot recover the record
You should drop and re-created the indexes on the
targret
3 if you want to use bulk mode on the target talb e
it will be used when you have more than 10L reocrods
4 in the source table
Normal
Data will laoded to the target table slower
we can recover the records

No need to drop and re-create

it will beuseful when source has less than
10L records
 DELETE TRUNCATE
1 DML DDL
2 Delete the rows from a table Delete all the rows from a table
3 we can give where we cannot where
4 Rollback no rollback
Truncate is always faster than delete
DROP
DDL
Delete the structure/ definiton of a table with the data permanenlt
we cannot give where
no rollback
 source qualifer lookup
1 Active passive
2 no cache index and data cache
3 connected connected /unconnected
4 only i/o I/O/L/R
5 ALL JOINES SUPPORTED BY SOURCONLY LEFT OUTER JOIN
6 syntax : select tabname.column1 , syntax : select table.column_name as clolumns
able.column_name as clolumns