[ceph-users] using secondary OSDs for reading

Hi, Quoting the page https://docs.ceph.com/en/latest/architecture/

location query over a chatty session. The CRUSH algorithm allows a client to compute where objects should be stored, and enables the client to contact the primary OSD to store or retrieve the objects.

So clients contact the *primary* OSD to store/retrieve objects. Why do clients not contact secondary (or even tertiary) OSDs for reading data? Would that not (potentially) result in greatly improved performance..? I'm sure there are good reasons for the current behaviour, but it seems logical, since you have multiple copies of the same data, to try and use the nearest copy? Curious :-) MJ