inventory data loading steps checklist

[Former Member]

hi experts ,

i am going to reload inventory data. after  study i have loading procedure . before going ahead with the procedure can any help me by correcting my loading process check list.

loading procedure steps:

step1:drop aggregates,stop process chain,delete data from infocube, delete data from psa

step2:take down time (00:00:01) ex:08.07.2012 (00:00:01)

step3:run v3 wit immediate,clear rsa7,deshedule v3 jobs by deleting released jobs

step4:delete data from setup table ,fill 2lis_03_bx setup table  mcnb

step5:load data to biw with full info package ,compress requst with marker update

step6:delete data from setup table ,fill setup table for 2lis_03_bf ,2lis_03_um  -----OLI1BW,olizbw(for last 1month from today )

step7:load data to biw with full info package(bf,um) and compress requst without  marker update

steps8:intialize 2lis_03_bf,2lis_03_um datasources

shall i release end users after step8 ???

should i cancel v3 job run while loading historical data  or shall i reschedule it ???

i assue that i am locking endusers while loading historical and going ahead with process

i wann load data from 2009 to till date .

step9:split data into different sets and will fill setup table for old data(bf,um)

  shall i run parallel loads for bf ,will go posting date wise??

step10:load data to bi with full infopackage and compress witout maker

step11:reshedule v3,job,rebuild aggregates,rerun infopackage .

please suggestion me if i miss out any step or doing in wrong way ???

thanks in advance .

regards,

hari

[Former Member · Accepted solution]

Glad you already read the guide 🙂

Here we go:

step3: run V3 again after a few minutes when the system is down.  Sometimes records are sent with a delay

step5: it's not a full load but 'generate initial status'.

step6: don't select on today but 31.12.9999 as it's possible to make postings in the future.  They are not extracted anymore by BF by delta as they already exist!  BX will extract these.  Unless you have a very good reason not to extract them.

You can release the users after step 8.  But don't run any delta or V3 job yet.  Just to keep an overview on things.

If it's possible, do this after historic movements are extracted.

Run in parallel as much as you can.  Speeds up things. 

In theory you can release the system after BX as historic movements should not change anymore.

For the rest, seems ok.

[Former Member · Accepted solution]

hi 2xu...

thank you soooo much for your reply ... u have helped in many situation ... thank u again

step3: run V3 again after a few minutes when the system is down.  Sometimes records are sent with a delay ...

  ya once  this happened to me ...

can you pls  expalin me step6 suggestion in details ....

regards ,

hari

[Former Member · Accepted solution]

In ECC one can do postings with a date in the future.  F.e. today a posting is made with posting date 01.08.2012.  When you run BX, this posting is taken into account for the initial stock balance (as in MB52).

Now, if you only extract the postings up to the date your are doing the new inititial load, this posting will be missing.  And with new delta extractions this posting won't be extracted either because it's already in the system and won't be written to the delta queue (in fact, it's a historic movement, even if the posting date is in the future).

With BX your stock balance will be correct, but this posting in the future won't be extracted.

If you want a 100% correct initial load, you should extract your historic movements with a posting date from ... to 31.12.2050 for example.

Now, a lot of companies don't do postings in the future, so maybe check in your company if they do this.  If not, then you can forget all this.  If they do, I'd extract them.

[Former Member]

Hi hari,

Please check infopackage , Go to tab -> scheduler ->  initilization option for source System -> if exit delete it.

Regards,

Anand Mehrotra.