cancel
Showing results for 
Search instead for 
Did you mean: 

PCL1-TE Cluster Structure USER/PTK99 leads to inconcistencies

Lukas_Weigelt
Active Contributor
0 Kudos

Hi guys,

System Information: ECC 6.0 EHP4 SP Level 46.

today I made a discovery which makes me feel uneasy. In the PCL1-TE Cluster, there is the Structure PTK99, respectively called "USER". This strucutre is designed for customer enhancements. I did use this possibility to add fields to the whole TRV-Process which were a requirement of our special department.

Everything worked out so far. After some time, I got additional requirements from the special department for more fields which I added per enhancement again.

Now here's what shocks me:

Whenever A Trip or Separation Allowance (PR05, TGANL, TGMOD, etc.) which has been created before the point of time of my alteration of PTK99 is tried to be accessed, changed, deleted, there is a dump. More detailed, whenever the standard macro "rp-imp-c1-te" is called and the trip/SA is older than my adjustment, it dumps for CONNE_IMPORT_WRONG_STRUCTURE, CX_SY_IMPORT_MISMATCH_ERROR.

The consequence I understand from this behaviour is: Whenever I change the cluster structure, old entries can't be accessed in any way anymore. This would be a catastrophy since alteration of PTK99 could also happen after GO-Live when requirements Change and it would corrupt all present entries.

What can be done to avoid this? What can be done to rectify this subsequently?

Best regards,

Lukas

Accepted Solutions (1)

Accepted Solutions (1)

0 Kudos

Hello Lukas

Hope your well! Yes, this is indeed a known issue   and it is quite a well known error that can occur after SP upgrade with custom appends because additional fields were added due to EU Directives regarding VAT refunds to various structures such as PTK99 PTK05, PTK03 (BELEG) etc and these are delivered with new support packages.

Delivering new fields causes no problems with old trips except if there has been an additional modification in the structure like in your case.


Once you modify a data cluster structure, then the order of the fields must then stay the same and the new fields must be added to the end of the structure so as SAP delivers new fields to the standard automatically the customer fields are moved to the end of the structure. For example:

Before the supportpackage:

Standardfield 1
Standardfield 2
Customer Append

After the supportpackage:

Standardfield 1
Standardfield 2
Standardfield 3 (new delivered field)

So to avoid this dump, the structure must now be sorted in this way:

Standardfield 1
Standardfield 2
Customer Append
Standardfield 3

Otherwise the dump appears because when reading the old trip the data does not fit any more to the order of the fields in the new structure definition!

One of my colleagues wrote a report that can re-sort the structure for you with the custom append re-inserted to avoid the dump and its the only way to avoid this scenario I'm afraid. If you want this sample report, please let me know.

Hope the info helps you out in the meantime!

All the best

Sally

Message was edited by: Sally Redmond

Lukas_Weigelt
Active Contributor
0 Kudos

Hey Sally!

I'm doing fine, thanks! Hope you're well too

Your explanation is interesting, since it doesn't 100% overlap with my problem but is still critical to be considered! I didn't know there can be problems through SPs like this as well. From your explanation I understand how it's technically working now. I'm subsequently veeeery interested in the report of your colleague you mentioned. In what kind of "forms" would you be able to send it to me? Transportable archives won't work out, since I'm employed in Public Sector and everything that comes "from outside" is considered vicious

Disregarding this, I'd like to elaborate on how I provoked the Cluster-Problem. Maybe the Report of your colleague won't even work here. Screenshot:

This is the Main include I appended to PTK99. What happens from time to time now while under development is, that I add fields of the Sub-Includes. I.e. the order of the fields within PTK99 itself changes.

For example:

PTK99

          Include A

                             Field A

                             Field B

                             Field C

          Include B

                              Field A

                              Field B

                              Field C

and so on...

And then, I add a field within one of the includes like

PTK99

          Include A

                             Field A

                             Field B

                             Field C

                             Field D (added)

          Include B <--- Cluster subsequently being influenced from here as I understood from your explanation.

                              Field A

                              Field B

                              Field C

Could this problem be avoided if I added fields with an approach like this?

PTK99

          Include A

                             Field A

                             Field B

                             Field C

          Include B

                              Field A

                              Field B

                              Field C

          Include C, added

                              Field A , added (would actually be Field D in Include A from a functional point of view but coudl just be added at the end to influence no further fields within PTK99, in theory).

Right now, in development, all that eventually happens is that we can toss away all our "old" test data once I change the PTK99 structure. The consequence is, that my colleagues are grumpy, because they have to maintain new test data , but nothing more.

My big fear, however, is, that we go live and after a few months or so I need to make changes on PTK99 again which will have an impact on the production, worst case scenario: all productive data can go into the litter bin. This must not happen 😕

Phew... well, well . Bottom line I'd really like to take a look at that rearrangement-report you offered.

Cheers! Lukas

0 Kudos

Hi Lukas

Certainly alot to consider there...

probably the report wont help therefore but at least it can give you an idea as to how you could re-sort the custom append within the new structure (this will be in the case of displaying OLD trips..new trips should be ok?)

Hope its useful!

Sally

0 Kudos

Hi Sally,

After patch upgrade, PTK17 structure change and I am facing the same issue, can you please share sample report to resort PCL1(TE) data.

Answers (0)