JOb Running for Hours

[vamsi.4a6]

Hi all,

Job is running for a long time


Orcale connector(source)-->aggregator--->transformer-->Orcale connector(Target)


source oracle connector consists of query using 4 tables joined based upon Join Keys.

Enabled partition=No, so the oracle connector is running in the sequentail Mode.
In aggregator 6 group by keys are present.Is there any way to improve the performance of Job by selecting Enabled partition=Yes or some other option?

[chandra.shekhar@tcs.com]

I think enabling Hash partitioning in the source would help.
And what about the partitioning at the target level ?
Also how much data we are talking about here ?

[vamsi.4a6]

@Thanks for the input

1)Target is key partition

2)I am running the JOb for 63 iterations using Loop activity in
Sequence.For each iteration the volume of input data is 50 Lakhs approximately

Clarification:1)Is select Enabled partition=YES and Hash partitioning will effect aggregator functionality?
2)Hash partitioning,oracle partitons,rowid are datastage partitions or oracle partitions?

[chandra.shekhar@tcs.com]

I think It will not affect the aggregator performance, as I have used the same in DB2. And for your 2nd point, hope this might help.

<corrected>

[chulett]

(copy-pasted from IBM Information center)

Please do not that. Link to articles in the Information Center, do not post their contents here wholesale.

[chandra.shekhar@tcs.com]

Aye Aye Captain..

[vamsi.4a6]

@chandra

Thanks for the input.one small clarification,If i am using multiple tables Joined based upon Join Keys or sub queries in source oracle connector I am not sure which table to use under option-Table name for partitioned reads?

[chandra.shekhar@tcs.com]

vamsi, I dont think that you'll be able to to do that.
If your query has multiple tables in it, then its difficult to achieve partitioned read.
But you can test it at your end by running the job and taking the driving table of your query for partitioned reads and match the count. Give it a try.

[vamsi.4a6]

If i am using multiple tables Joined based upon Join Keys or sub queries in source oracle connector I am not sure which table to use under option-Table name for partitioned reads?

Could anybody please clarify this doubt?

[chulett]

You. As noted. Assuming your tables are actually partitioned.