[Bug 483834] [USS] Should INDEX requests make a snapshot and then operate on a cursor of that snapshot?
|Reported||Dec 07, 2015 13:30 EDT|
|Modified||Nov 08, 2016 11:28 EDT|
Iterating over the blobs of an application involves multiple requests with "pagesize" and "page" arguments. These arguments kind of expect that the no blobs are added or removed for the duration of the iteration.
I opened this bug to discuss whether that's just accepted or whether we want a more reliable approach to blob iteration...