On Fri, 8 Jun 2012, Martin Wilderoth wrote: > Hello and thanks, > > I have removed the data and metadata pool. Do I need to create them again, or > will they be created automaticly. > Maybe I need the undocumented way of creating the mds map ?. I would like to > get an empty cephfs to play with again. Kids, don't try this at home. 1- Recreate the pools with something like ceph osd pool create data PGNUM ceph osd pool create metadata PGNUM where PGNUM is something like 50*num_osds, maybe rounded to a power of two. 2- get the numeric pool ids for data and metadata (the number after "pool") ceph osd dump | grep ^pool 3- create a new mdsmap with those pools ceph mds newfs METADATAPOOLID DATAPOOLID WARNING: This will clobber any fs data that is in those two pools. And make any data in the previous mdsmap's pool inaccessible. In fact, I'm going to add a --yes-i-really-mean-it argument to this command now. That should do it! sage > > /Best Regards Martin > > > On Fri, 8 Jun 2012, Tommi Virtanen wrote: > > > On Fri, Jun 8, 2012 at 12:35 AM, Martin Wilderoth > > > <martin.wilderoth@xxxxxxxxxx> wrote: > > > > And if i just want to mkfs or similar in mds is that possible. Or or my rbd will > > > > also be removed ? Or how should I recover if I don't care of the data in > > > > the POSIX file system ? > > > > > > By default, the Ceph Distributed File System puts file metadata in the > > > pool "metadata", and file contents in the pool "data". RBD images, by > > > default, are stored in the pool "rbd". This means you can empty/remove > > > the pools "metadata" and "data" without affecting RBD. > > > > There is also an undocumented way to recreate a new mdsmap (mds cluster) > > without affecting the rest of cluster. Is that what you're after? > > > > sage > -- > 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 > > -- 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
