create index on table with blobs

Hi all, perhaps there is anyone here who could help me to find a solution for this problem. In an ODS 7.23UC4 we have defined a table with the following colums: mandt char(3) relid char(2) srtfd char(30) srtf2 decimal(10) clustr decimal(5) clustd byte The table is fragmented by round robin, 16 x 2 GB alloc.; currently it has about 80.000 rows and it has an index on (mandt,relid,srtfd,srtf2). As you can see the field clustd is declared as byte, so this table will contain blobspace (but the table is located in a normal dbspace). Now comes our problem: When we try to create the index on this table using PDQ functionality, the process doesn't work parallel at all. It does a sequential table scan, each fragment alone. The time to create the index is about 16 min. But: When we create a second table with the same columns and copy all the data from the first to the second table, but not the data in the clustd byte ("insert into tablenew(mandt,relid,srtfd,srtf2,clustr) select mandt,relid,srtfd,srtf2,clustr from tableold"), and if we create the same index on the second table, then all works fine. The fragments will be scanned in full parallel access, and the index will be created in about 10 seconds. (By the way, PSORT_NPROCS, PDQPRIORITY and all other important environment variables are set correctly.) So, can it be, that the PDQ functionality doesn't work for a table containing blobs? Or, if someone here knows how it works (special parameter, environment variables,etc.) please give me a tip. Please answer via email to rolandw@wartenberg.mannheim-netz.de. Regards, Roland