Maximising performance of time-based query

Greetings gentlemen,

I'm embarking on a project to migrate all data from a Centera cluster to A.N.Other storage medium, the first stage of which is simply to get a list of all C-Clips on the cluster (starting at time zero). I'm keen to do this in as efficient a manner as possible, especially after reading the API docs which suggest that typically I could expect 100 results a second, which for 100 million C-Clips would take about 10 days.

So my question is simply this - if I just format my FPQueryExpression using setStartTime(0) and setType(FP_QUERY_TYPE_EXISTING), is this more efficient than if I also select particular fields using selectField()?

It seems possible to me that using selectField() might force the SDK to actually read the CDF into memory. Or does it do that anyway?

Any help appreciated,

Geoff.