Previous Page | Next Page

Creating and Loading SPD Engine Files

SPD Engine Component File Naming Conventions

When you create an SPD Engine data set, many component files can also be created. SPD Engine component files are stored with the following naming conventions:

filename.mdf.0.p#.v#.spds9
filename.dpf.fuid.p#.v#.spds9
filename.idxsuffix.fuid.p#.v#.spds9
filename.hbxsuffix.fuid.p#.v#.spds9

filename

valid SAS filename.

mdf

identifies the metadata component file.

dpf

identifies the partitioned data component files.

p#

is the partition number.

v#

is the version number. (footnote 1)

fuid

is the unique file ID, which is a hexadecimal equivalent of the primary (metadata) path.

idxsuffix

identifies the segmented view of an index, where suffix is the name of the index.

hbxsuffix

identifies the global view of an index, where suffix is the name of the index.

spds9

denotes a SAS 9 SPD Engine component file.

Table 2.2 shows the data set component files that are created when you use this LIBNAME statement and DATA step:

libname sample spde '/DATA01/mydir'
   datapath=('/DATA01/mydir' '/DATA02/mydir')
   indexpath=('/IDX1/mydir');
data sample.mine(index=(ssn));
   do i=1 to 100000;
   ssn=ranuni(0);
   end;
run;

Data Set Component Files
mine.mdf.0.0.0.spds9 metadata component file
mine.dpf.000032a6.0.1.spds9 data file partition #1
mine.dpf.000032a6.1.1.spds9 data file partition #2
mine.dpf.000032a6.n-1.1.spds9 data file partition #n
mine.dpf.000032a6.n.1.spds9 data file partition #n+1
mine.hbxssn.000032a6.0.1.spds9 global index data set for variable SSN
mine.idxssn.000032a6.0.1.spds9 segmented index data set for variable SSN


FOOTNOTE 1:   The version number increases only when the data set is updated, that is, when the data set is opened in UPDATE mode. Operations such as PROC SORT that replace the data set reset the version number to one, instead of incrementing it. [arrow]

Previous Page | Next Page | Top of Page