Index Seperator Character

[Former Member]

Hi Experts,

on our 10.2.0g Oracle Database I observed that the "Delimiter" Character on some (but not all) Indexes are changed from tilde (~) circumflex (^) - Does anybody have a clue what is the background for this change?

We are running SAP 6.40 and 7.00 on HP Itanium on Oracle 10.2.0 - all System are NON-Unicode

Kind regards

Alex

[lbreddemann · Accepted solution]

Hi there!

The name of the index has nothing to do with the database (version) you're using.

It's created by the ABAP DDIC database interface layer.

Check function:

FUNCTION DD_INDEX_NAME.
*"----------------------------------------------------------------------
*"*"Lokale Schnittstelle:
*"       IMPORTING
*"             VALUE(INDEXNAME) LIKE  DD12L-INDEXNAME
*"             VALUE(TABNAME) LIKE  DD02L-TABNAME
*"             VALUE(DBSYS) LIKE  SY-DBSYS DEFAULT SY-DBSYS
*"       EXPORTING
*"             VALUE(DBINDEX) LIKE  DD12L-DBINDEX
*"----------------------------------------------------------------------
* Namenskonventionen für Indizes:
* Ab Release 4.0: (Siehe Anmerkung)
*   Tabname+Divider+Indexname
*   Als Divider ist normalerweise die Tilde (~) vorgesehen.
*   Für das Austauschverfahren wird aber ein zweiter Name benötigt.
*   (Index auf Schattentabelle soll beim Repository-Switch, also
*    beim Rename der Tabelle, bereits einen gültigen Namen haben).
* Release 3.x:
*   Tabname(10)+Indexname
*   Tabname wird auf 10 Stellen mit _ aufgefüllt, wenn er kürzer ist.
*   Indexname steht linksbündig ab der 11ten Stelle.
* Release 3.x für Austauschtabellen
*   Tabname(10)+Indexname(3)+'X'.
*   (Begündung siehe oben.) Der Indexname wird mit _ auf drei Stellen
*   aufgefüllt. Dann folgt das Zeichen 'X'.
* Release < 3.0
*   Tabname(7)+Indexname(1).
*   Vor Release 3.0 war der Indexname nur eine Stelle lang und
*   Tabellennamen waren in den ersten 7 Stellen signifikant.
*   Kürzere Tabellennamen wurden mit _ aufgefüllt.
  DATA RC LIKE SY-SUBRC.

Although the comment is in german, I guess babelfish will allow to understand it...

regards,

Lars

[Former Member · Accepted solution]

Hi Lars,

thanks for your posting - which absolutely explains the behavior to me.

btw. as I'm German native speaker I avoided to use Babelfish whose translation might make me confused

Kind Regards

Alex

[former_member213250]

Hi Alex

What type of indexes you are having this delimeter change

primary index or secondary indexes

Did you tried to rebuild/recreat these index and see if the same delimeter ' ^ ' instead of ' ~ ' showing up.

Regards

Venkat

[former_member213250]

Hi Alex

What type of indexes you are having this delimeter change

primary index or secondary indexes

Did you tried to rebuild/recreat these index and see if the same delimeter ' ^ ' instead of ' ~ ' showing up.

Regards

Venkat

[Former Member]

Hi Venkat,

Thanks for your response

the related indexes are primary as well as secondaries - e.g.

SQL> select index_name from dba_indexes where index_name like 'TSTC%';

INDEX_NAME

-

---

TSTC_SRT^0

TSTCA^0

TSTCA_C~0

TSTCC~0

TSTCP^0

TSTCRID~0

TSTCT^0

TSTC^0

TSTC^001

TSTC_LOG~0

TSTC_LOG~1

Yes, in the past we did some rebuilds of indexes on most of our systems ... As I recognized this topic today for the first time, I unfortunately do not know if the Deleimiter-Change is related to the rebuild.

Kind regards

Alex

[former_member213250]

Hi Alex.

I don't think that rebuild of index in previous has and effect of changing delimeter .

I have came across such kind of issue in previous but that was regards to BW index.

I don't think this change will have any impact on system performance.

So for now can you try to rebuild one of these index and see whether index representation

is with ' ~ ' or with ' ^ '

Regards

Venkat