It's worth noting that the aggregate metadata is user-facing in a sense that there is an API for showing aggregate details, including the aggregate metadata [1]. The primary consumer of the metadata is Nova itself, but it's entirely possible for an end user to write code to create/update/read aggregate metadata. And I get wary of the idea of put data != get data for the end user.
It's worth noting that the aggregate metadata is user-facing in a sense that there is an API for showing aggregate details, including the aggregate metadata [1]. The primary consumer of the metadata is Nova itself, but it's entirely possible for an end user to write code to create/update/read aggregate metadata. And I get wary of the idea of put data != get data for the end user.
[1] http:// developer. openstack. org/api- ref-compute- v2.1.html# showAggregate