|
This is very likely the way that it works. Recreating the collection
is not the same as recreating a test. The test is the basic unit of
DTM and once you create one, you ought to consider it sacrosanct.
The collection on the other hand is the current job or task that DTM
has been assigned to do. As part of the initialization of the task,
DTM 'creates' the collection. As part of that initialization, a lot of
things are resolved. Unlike other computing tasks, the collection does
not go away when completed so the init is 'sticky' in that sense. You
may want to review the results many times until you resolve all the
problems with the run. If there are no changes such as what you are
doing, you can always rerun the collection.
It's similar to changing something in sysgen on VMS. You need to
reboot in order for the change to take effect.
As far as the docs go, I bet it's hidden in there somewhere. If not,
you jot it down in your DTM things learned by trial and error.
|
| SET BENCHMARK_DIRECTORY will instructs DTM to establish the
[new_benchmark_directory] as the default directory for benchmark files of the
new creating collection, however, this change won't be reflected in the
existing collection unless you recreate it using the RECREATE command.
It is also true if you modify test descriptions or the contents of groups
after including them in a collection, those changes also are not reflected for
that collection unless you recreate it.
This is because in the process of creating a collection, DTM scans
all of the tests to be added to the collection and stores their attributes
within the collection. For any tests that do not contain an explicit
benchmark specification, the default benchmark location is applied at that
time and stored in the collection. Subsequent changes to the default
benchmark location will have no effect on the collection unless it is
recreated.
We agree the documentation could be more clear about this and will raise it as
a documentation wish for DTM.
Fred, thanks.
-James
|