I have written the below relevance to fetch the count of filesets with specific version level.
This relevance takes a lot of time in evaluation.
Do let me know how can I optimize this relevance.
Q: (number of (latest filesets ((unique values of lpp_names of filesets whose (version of it = (((character 0 of it) & "." & (character 1 of it) & "." & (character 6 of it)) of (first 7 of (maxima of current service packs of operating system as string)))) of object repository) as string) of object repository))
A: 384
T: 680716
Seems like AIX, which I don’t have a test box handy, so my suggestion will not be a tested one.
You have the object repository in there twice. If you have 10 objects, and invoke this way, you will end with 10 x 10 instances in cost or 100 instances, instead of the 10.
Can you test these for speed and the expected 384 number of results?
number of latest filesets (filesets of object repository) of object repository
This one avoids the latest fileset (string) inspector , but I don’t know if it will meet your use case, as I am not sure if it is possible to install a fileset with version that is greater than the service pack version, which would be filtered out by the latest fileset (string) method.
number of lpp_names of items 1 of ((((character 0 of it) & "." & (character 1 of it) & "." & (character 6 of it)) of (first 7 of (maxima of current service packs of operating system as string))) ,(filesets of object repository)) whose (item 0 of it = version of item 1 of it)
This one keeps the latest fileset (string) inspector but I think it will still be expensive.
number of ((((character 0 of it) & "." & (character 1 of it) & "." & (character 6 of it)) of (first 7 of (maxima of current service packs of operating system as string))),(latest filesets (filesets of object repository) of object repository)) whose (item 0 of it = version of item 1 of it)
