Is there a way to configure Profiler so that it understands zpools on Solaris i.e recognizes that the total size of each file system in zpool shares the total space and doesn't report them all cumulatively?
I know Profiler sort of supports ZFS file systems - if you edit the mod.sys.strg.Storage.xml file on a Solaris client you can edit the strings in the valid and invalid file system types fields so that Profiler will see them as valid filesystems, however if you have multiple ZFS file systems in a pool (a very common configuration) it reports that server as having more disk space than it actually has.
For example, we've just built a Solaris server which has a 100TB zpool. There are 14 separate ZFS file systems in this pool. If I add each file system into the FSDevicePattern field, Profiler will report 14 x 100TB file systems (with a total of about 60TB of data currently), which as you can imagine, would completely distort my reports (and make my storage utilization look pretty bad!) If I add just one valid file system (the largest single one) it would show that server with only, say 10TB used on a 100TB file system which would also not be much use.
Profiler does not support zPools at this time. We are interested in supporting it in the future, but it has not made it to the roadmap yet.