Hi, Looking at the reporting from command line tools (ceph status, pg query, etc), I don't find a way to tell how degraded the recovering objects are for a given PG. The use case could be, when doing deployment, we don't want to stop the procedure upon a recovering PG, but at the same time, we don't want to stop an OSD which could bring *unfound* objects (or even stuck at peering PG). Any idea how we could query those information? Or should we expose those information from CLI, some ways I can think of: 1. Implement a command line option like "ceph osd could-down <ids> [<ids>...]" which would examine if the down of the given OSD(s) would bring problems? 2. Instead of just warning for "degraded", we also warn if a PG could further lose OSDs, so that we stop upon having that warning? Thanks, Guang -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
