]> git.apps.os.sepia.ceph.com Git - ceph.git/commit
tasks/ceph: conservative PG count in FS pools
authorJohn Spray <jspray@redhat.com>
Tue, 2 Dec 2014 13:30:06 +0000 (13:30 +0000)
committerJohn Spray <jspray@redhat.com>
Tue, 2 Dec 2014 13:30:06 +0000 (13:30 +0000)
commit89beed1b0a46d1e55eb9d65f8144092d753a1fb2
treee232f4d31ab110adb392d172ae451cce1dc5e66f
parentfd3aaec8c3542668754e7d7ce8163a5a8d976d8d
tasks/ceph: conservative PG count in FS pools

This was tripping over the recent commit 42c85e80
in Ceph master, which tightens the limits on
acceptable PG counts per OSD, and was making
teuthology runs fail due to never going clean.

Rather than put in a new hardcoded count, infer
it from config.  Move some code around so that
the ceph task can get at a Filesystem object
to use in FS setup (this already has conf-getting
methods).

Signed-off-by: John Spray <john.spray@redhat.com>
tasks/ceph.py
tasks/ceph_manager.py
tasks/cephfs/filesystem.py