DTP Process!

[Former Member]

Hello Gurus,

Need suggestions on this - (I searched in the forums before posting this but could not find answers to my questions.)

I know how DTP works and how to deal with error stack but I want to know how it behaves?

1.What happens if I am doing a full load (1st Day) and some records came into error stack. 2nd Day, again my full load started and I have not cleared my error stack of 1st Day and some more records from 2nd Day came into my error stack? What needs to be resolved first before my 3rd Day full load starts or next full load starts? How does DTP behaves on this? Will it recognize my Day1, Day2 error records separately when I correct them or will it push all of then at a time.

What steps needs to be followed before my next full load starts? How to resolve this problem?

2. What happens if I am doing two subsequent delta loads into one data target and my first process is not yet finished? I know that delta capabilities in DTP will load both of them but if both the delta request has the same keys for update, will my load fails? or will it be successful?

How does DTP behaves and how do we need to handle this kind of situation?

3. Is it compulsory in both the cases that either full or delta but first set of request or load needs to be activated before next one processed or not necessarily required?

Thanks in Advance,

Liza.

[Former Member · Accepted solution]

I put my thoughts in here..you may wait for Gurus for their comments

<b>1.What happens if I am doing a full load (1st Day) and some records came into error stack. 2nd Day, again my full load started and I have not cleared my error stack of 1st Day and some more records from 2nd Day came into my error stack? What needs to be resolved first before my 3rd Day full load starts or next full load starts? How does DTP behaves on this? Will it recognize my Day1, Day2 error records separately when I correct them or will it push all of then at a time.</b>

I presume all your errors will summed up

Day 1- 5 erroneous records and you didnt correct

Day 2- 5 (from Day 1) + 2 NEW erroneous records and and you didnt correct

Day 3 - 5 (from Day 1) + 2(from Day 2) + 3 NEW erroneous records and and you didnt correct

total on day 3 you should see 10 erroneous records which you need to correct and push forward.

In principle all your erroneous records should get accumulated as you move ahead.

<b>

What steps needs to be followed before my next full load starts? How to resolve this problem?</b>

<b>2. What happens if I am doing two subsequent delta loads into one data target and my first process is not yet finished? I know that delta capabilities in DTP will load both of them but if both the delta request has the same keys for update, will my load fails? or will it be successful?</b>

Technically you can schedule parallel DTP's for same data targets ;however you have to make sure you have enough parallel

process in your system to handle them and so as not to fail .

In case of data target as DSO i reminisce discrepancies of data in ACTIVE table when loaded in parallel;however it was 3.1C and can say more of exception than rule and with nw2004s am sure DSO matured pretty much by now.

Infocubes should have no problems with parallel DTP'S.

PS - if you are sp< 11 you might get locking issues with DTP when scheduled in parallel.

Note 998093 - P11:DTP:PC:Parallel DTPs that update the same data target

for write optimized DSO you can do paralle DTP's with SP 12

Note 1007769 - Parallel updating in write-optimized DataStore objects

<b>How does DTP behaves and how do we need to handle this kind of situation?</b>

3. Is it compulsory in both the cases that either full or delta but first set of request or load needs to be activated before next one processed or not necessarily required?

Its not Imperative as aforementioned.

Hope it Helps

Chetan

@CP..

[Former Member · Accepted solution]

Thanks so much Chetan and most of my doubts cleared and can you clarify this - If I have 2 delta request with the same keys for update, will my load fails? or will it be successful?

Thanks a lot!

Liza.

[Former Member · Accepted solution]

welcome ..glad it helped

Load would be Successful .

Hope it Helps

Chetan

@CP..

[Former Member · Accepted solution]

One again thank you so much!

Liza.

[Former Member · Accepted solution]

welcome

Chetan

@CP..

[former_member192700 · Accepted solution]

Hi together,

I have to correct Chetan's answer on the first question as follows:

1.What happens if I am doing a full load (1st Day) and some records came into error stack. 2nd Day, again my full load started and I have not cleared my error stack of 1st Day and some more records from 2nd Day came into my error stack? What needs to be resolved first before my 3rd Day full load starts or next full load starts? How does DTP behaves on this? Will it recognize my Day1, Day2 error records separately when I correct them or will it push all of then at a time.

<b>In general, some comments upfront. It depends mainly on the settings for the error handling that you set in the DTP. All the discussion is anyway only relevant in case that you set it to 'valid records update , reporting possible (GREEN)'. If you set the error handling to 'valid records update, no reporting (RED)', you have for any further DTP loads first to get ride of the RED request. You can do so by manually setting it to green, delete it or repair it (the latter two options will be provided by the system on the attempt to execute a DTP for a target that has already a RED request).</b>

I presume all your errors will summed up

<b>==> the errors are not summed up, yet on every request you will get a new request number for the error stack</b>

Day 1- 5 erroneous records and you didnt correct

Day 2- 5 (from Day 1) + 2 NEW erroneous records and and you didnt correct

<b>==> 2 error stacks requests, one per load (1 contains 5 records the other 2 records)</b>

Day 3 - 5 (from Day 1) + 2(from Day 2) + 3 NEW erroneous records and and you didnt correct

<b>==> same as above + a third error stack request with 3 records</b>

total on day 3 you should see 10 erroneous records which you need to correct and push forward.

<b>==> you have 3 error stack request which you can process (after correction) with one error DTP (can be created from the DTP maintenance of the DTP causing the error)</b>

In principle all your erroneous records should get accumulated as you move ahead.

Cheers

SAP NetWeaver BI Organisation

[Former Member · Accepted solution]

danke soviel Andreas..

thanks for correcting my thoughts and putting a real case example

You see Liz..i mentioned earlier to wait for Guru's - (Andreas)

Haben Sie einen Nizza Tag!!!

Chetan

@CP..

[Former Member · Accepted solution]

Thanks to both for clearing by doubt.

Thanks,

Liza.