> > On fsY (ceph) you might see > > > > chunk_bytes=65536 > > stripe_count=32 > > mirror_count=3 (ceph ignores) > > raid_type=1+0 (ceph ignores) > > max_object_size=64MB (ceph adds) > > Hmm, I thought Ceph had replication for files? "mirror_count" was intended > to indicate the number of copies of that file that are maintained. Maybe it > needs a better name? Yeah, it was a weak example. Ceph does replication, but it's not exactly specified that way: files are allocated to a storage "pool", which has a particular replication level. A pool's replication level can be adjusted on the fly and the object storage cluster will magically 'make it so'. The pool's mapping rule can also be set such that, for example, objects are allocated to a subset of drives (say, a sub cluster of big+slow SATA boxes). I haven't settled on a nice way to expose to file -> pool -> object_layout mapping to users, so this is a bit up in the air. The striping onto objects, at least, is straightforward :). > Even dynamically-added "comments" might be OK, like: > > chunk_bytes=65536 # default > stripe_count=32 # default > mirror_count=3 > raid_type=1+0 # default > max_object_size=64MB # unknown I like this the best, because it's pretty clear how both user and fs should interpret the comments. sage -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
