SPLITNCV with wildcard

Dear All,

one test is generating the following data for example:

S3ESNX0J951635M-read_data_units : 21504913
S3ESNX0J951635M-written_data_units : 70723823
S3ESNX0J951635M-avail_spare : 100
S3ESNX0J951635M-percent_used : 66
S3ESNX0J951626N-read_data_units : 20199161
S3ESNX0J951626N-written_data_units : 68858853
S3ESNX0J951626N-avail_spare : 100
S3ESNX0J951626N-percent_used : 66
2J4520102682-read_data_units : 1199257
2J4520102682-written_data_units : 4055284
2J4520102682-percent_used : 0
2J4520102682-avail_spare : 100

I want *data_units* to be gauge, while the others should remain derive.
What does the SPLITNCV line in xymonserver.cfg have to look like? I've
tried the following without success:

SPLITNCV_ssd="data_units:GAUGE"
SPLITNCV_ssd="*data_units:GAUGE"
SPLITNCV_ssd=".*data_units:GAUGE"
SPLITNCV_ssd="%.*data_units:GAUGE"

Thank you for your help!

Regards