RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2024-12-25 21:03:31 +00:00
Code Issues Packages Projects Releases Wiki Activity
5,183 Commits 211 Branches 394 Tags 1.3 GiB
568c25de03
Commit Graph

1409 Commits

Author SHA1 Message Date
Loic Dachary
a7194e5685 openstack: thrash_cache_writeback_proxy_none.yaml need bigger disks 
http://tracker.ceph.com/issues/15324 Fixes: #15324

Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-30 18:05:08 +02:00
Sage Weil
935a7eb229 rados/singleton-nomsgr/all/full-tiering: s/readforwad/readproxy/ 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-30 08:23:31 -04:00
Sage Weil
d823ff6864 rados/singleton: thrash proxy (not forward) cache mode 
See #12814

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-30 08:23:31 -04:00
Loic Dachary
d9a8f49fc5 openstack: define the resources required to run a rados suite 
Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-30 01:45:14 +02:00
Josh Durgin
c2033806d6 Merge pull request #902 from smithfarm/wip-14984 
rbd: add singleton to assert no rbdmap regression

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2016-03-29 07:43:15 -07:00
Loic Dachary
c2ac83fa71 Merge pull request #907 from dachary/wip-15236-infernalis-x 
upgrade/infernalis-x: allocate OpenStack volumes for each OSD

Reviewed-by: Nathan Cutler <ncutler@suse.cz>
 2016-03-29 15:22:15 +02:00
Sage Weil
d303c0db75 Merge remote-tracking branch 'gh/jewel' 2016-03-29 08:46:04 -04:00
Loic Dachary
c6bf2cd4a2 Merge pull request #908 from dachary/wip-archs 
archs: add files for each known architecture
 2016-03-29 01:01:32 +02:00
Josh Durgin
7e595e1038 Merge pull request #869 from ceph/wip-15093-jewel 
Added blogbench.yaml to parallel before and after -x upgrade steps

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2016-03-28 15:57:03 -07:00
Yuri Weinstein
6122002cde Added blogbench.yaml to parallel before and after -x upgrade steps 
Fixes #15093

Signed-off-by: Yuri Weinstein <yweinste@redhat.com>
 2016-03-28 15:44:41 -07:00
Gregory Farnum
e541123998 Merge pull request #827 from ceph/jewel-14805 
Disable client-side permission checking in Hadoop.

It's not set up to pass appropriate UIDs/GIDs through, so everything breaks.
 2016-03-28 13:47:48 -07:00
Loic Dachary
d4e3cec185 dummy: reduce run time, run user.yaml playbook 
Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-28 18:30:06 +02:00
Gregory Farnum
2f5c76cf56 Merge pull request #898 from ceph/wip-15206 
test_msgr: improve debug ms level
 2016-03-28 09:24:42 -07:00
Loic Dachary
a8f4321f5c buildpackages: build will work on a range of flavors 
When selecting a flavor to build, we prefer the largest we can find and
fallback to smaller ones, down to the minimum acceptable flavor, if
there is no other available.

Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-28 18:15:12 +02:00
Loic Dachary
40277ca6a9 archs: add files for each known architecture 
So that they can be prepend to teuthology-suite to run a job on a
designated architecture.

Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-28 18:15:12 +02:00
Yuri Weinstein
0cb6468a4a Merge pull request #910 from ceph/wip-oldupgrades 
suites/upgrades: clean up old upgrade suites
 2016-03-26 08:04:21 -07:00
Sage Weil
843619d3b0 suites/upgrade: remove stable upgrade suites 
These are maintained in their respective branches (e.g.,
the latest suites/upgrade/hammer is in the hammer branch).

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-26 09:22:51 -04:00
Sage Weil
1686a202ea suites/upgrade: remove obsolete *-x suites 
These versions can no longer upgrade to jewel.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-26 09:22:09 -04:00
Sage Weil
2b3ec8de8d suites: run rados bench for max 180 seconds 
If we want a longer runtime than that, run it several times in
sequence.

This limits the amount of disk space we'll consume.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-26 09:18:41 -04:00
Greg Farnum
2e10639e56 fs: remove extra ceph-fuse invocation from TestMisc yaml fragment 
If we ever re-enable the kernel client on this subsuite (unlikely) we'll
need to move it, but for now that yaml frag belongs in the mount/ dir.

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
 2016-03-24 17:47:57 -07:00
Yan, Zheng
b977e86f5d tasks/cephfs: test for request caps during lookup/open 
Signed-off-by: Yan, Zheng <zyan@redhat.com>
 2016-03-24 17:47:53 -07:00
Loic Dachary
dc0b181299 upgrade/infernalis-x: allocate OpenStack volumes for each OSD 
The intention is to run with xfs which requires a separate disk for each
OSD. In addition the tests will exhaust the available root disk space
otherwise.

http://tracker.ceph.com/issues/15236 Fixes: #15236

Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-24 20:40:42 +01:00
Sage Weil
fcde1f6b9f suites/upgrade/hammer-x: do upgrade before staritng working 
Otherwise crushtool test will fail while creating pools.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-23 20:30:30 -04:00
Gregory Farnum
1a12d9cae6 Merge pull request #853 from ceph/wip-cephfs-test-python 
suites: invoke libcephfs python tests
 2016-03-23 17:06:40 -07:00
Jason Dillaman
08b6b2c415 Merge pull request #904 from dachary/wip-rbd-upgrade-use-jewel 
upgrade: use infernalis instead of hammer for rbd tests

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-03-21 16:20:14 -04:00
Loic Dachary
fb8c720bfc upgrade: use infernalis instead of hammer for rbd tests 
There are only few differences at the moment but it may be confusing
later on if a bug is fixed in the infernalis workunits later on.

Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-21 20:58:02 +01:00
Jason Dillaman
df622bd0c0 Merge pull request #895 from ceph/wip-rbd-mirror-test 
suites/rbd: small cleanup and add ceph_test_rbd_mirror workloads

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2016-03-21 09:06:09 -04:00
Sage Weil
1b6795a716 upgrade/infernalis-x/point-to-point: do upgrade before test workload 
Test creates pools, that can fail while crushtool is being
overwritten.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-21 07:23:51 -04:00
Sage Weil
83f1f46e4a upgrade/hammer-x/f-h-x-offline: fix mon statup 
OSDs are down, so we can't wait for healthy.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-21 07:10:45 -04:00
Josh Durgin
8e53f38618 suites/rbd: add ceph_test_rbd_mirror workloads 
Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-03-20 23:57:19 -07:00
Nathan Cutler
c4237debdd rbd: add singleton to assert no rbdmap regression 
http://tracker.ceph.com/issues/14984 References: #14984

Signed-off-by: Nathan Cutler <ncutler@suse.com>
 2016-03-20 21:50:53 +01:00
Sage Weil
eb8c02ab72 upgrade/infernalis-x/point-to-point: run rgw + s3tests in sequential 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 14:27:57 -04:00
Sage Weil
6111939365 upgrade/infernalis-x/parallel: do upgrade before workload 
This avoids upgrades while pools are being created (which may run
crushtool and fail).

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 14:24:57 -04:00
Sage Weil
534e39ac1d upgrade/hammer-x/f-h-x-offline: stop mons during upgrade 
Mons also can't go firefly -> hammer and form a quorum; stop them
first.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 14:15:24 -04:00
Sage Weil
917bc9ac02 upgrade/infernalis-x/point-to-point: upgrade before workload 
Upgrading the mons can make pool creation fail.  Don't do it in
parallel with the rados api test workload (which does a lot of
that).

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 09:21:22 -04:00
Sage Weil
e58ef3cbc0 upgrade/infernalis-x/point-to-point: run steps in series 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 09:20:45 -04:00
Sage Weil
4ffaf2ba5e upgrade/infernalis-x/point-to-point: remove stray upgrade steps 
This looks like an accidental cut and paste.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 09:20:30 -04:00
Sage Weil
a7e9d16c74 upgrade/infernalis-x/parallel: do workload in series 
Same fix as b483793f67.  This makes
the cls test run later so that it doesn't race with the package
upgrade.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-20 09:14:57 -04:00
Haomai Wang
322b79fd26 test_msgr: improve debug ms level 2016-03-20 11:54:49 +08:00
Loic Dachary
1cc657798a buildpackages: provision a small build machine for tests 
Signed-off-by: Loic Dachary <loic@dachary.org>
 2016-03-19 11:15:22 +01:00
Josh Durgin
f3aeb7bd47 suites/rbd/maintenance: use symlinks for cluster config 
This is the same 3-node cluster as most tests use, and qemu with
xfstests already has an openstack config in the qemu sub-suite.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-03-18 17:41:41 -07:00
Josh Durgin
6429e2b703 Merge pull request #893 from ceph/wip-rbd-maint-ops 
suite/rbd: test dynamic features and rebuild object map

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2016-03-18 14:23:23 -07:00
Sage Weil
b483793f67 upgrade/hammer-x/parallel: run workloads in sequence, not parallel 
Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-18 16:47:54 -04:00
Jason Dillaman
e463ac2cd1 suite/rbd: test dynamic features and rebuild object map 
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
 2016-03-18 14:17:34 -04:00
Samuel Just
3faabb80ac Merge pull request #888 from ceph/wip-15171 
suites/rados/thrash: vary min_{read,write}_recency_for_promote

Reviewed-by: Samuel Just <sjust@redhat.com>
 2016-03-18 10:58:15 -07:00
Josh Durgin
57766fdcb8 Merge pull request #892 from ceph/wip-rbd-qemu-journaling 
suites/rbd: add qemu tests with journaling

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2016-03-18 08:56:27 -07:00
Jason Dillaman
5caa4ed6b1 suites/rbd: add qemu tests with journaling 
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
 2016-03-18 11:35:56 -04:00
Yuri Weinstein
6aad8c07f7 Changed run time for radosbench to 200 
Signed-off-by: Yuri Weinstein <yweinste@redhat.com>
(cherry picked from commit 0ad1e4006b)
 2016-03-17 08:10:05 -07:00
Sage Weil
7567c80885 Merge remote-tracking branch 'gh/jewel' 2016-03-17 09:27:07 -04:00
Sage Weil
eb4b43b137 upgrade/hammer-x: fix ec portion of workload 
This should go in the workload sequence.

*also*, it'll be the first item, which pushes the many pool
creations in the api tests off until after packages install,
which avoids a problem with ceph-mon running crushtool while it
is being extracted.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-17 09:26:57 -04:00