Sage Weil
2a08cbbed5
qa/tasks/thrashosds,ceph_manager: thrash pg_remap[_items]
...
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:10 -04:00
Sage Weil
bd49dc0863
test/cli/osdmaptool/remap.t: add --remap test
...
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:09 -04:00
Sage Weil
ae557d0557
test/cli/osdmaptool: fix tests
...
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:09 -04:00
Sage Weil
d40825377e
osdmaptool: add --remap to run the remap_pgs() method
...
Output ceph cli commands to realize the new remaps entries.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:09 -04:00
Sage Weil
0364b3f173
osd/OSDMap: clean_pg_remaps
...
Helper to clean out no-op pg_remap cruft:
- pg_remap entries that match the raw result
- pg_remap_items entries where the 'from' part of the (from,to)
set isn't present.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:09 -04:00
Sage Weil
7f1934cb1e
osd/OSDMap: implement remap_pgs
...
Run a specified number of iterations trying to install new
pg_remap_items entries that improve the PG distribution.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:09 -04:00
Sage Weil
06ec9d41eb
crush: implement try_remap_rule
...
Simulate a CRUSH mapping but try to identify alternative OSD
choices (based on an underfull list and overfull set) that still
respect the CRUSH rule constraints.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-28 10:12:09 -04:00
Sage Weil
e110ffde50
crush: make get_immediate_parent const
...
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-25 14:53:14 -04:00
Sage Weil
a6aa42f793
mon/OSDMonitor: add 'osd [rm-]pg-remap[-items] ...' commands
...
Add commands to add and remove pg_remap mappings.
Require that a mon config option is set before this is allowed so that
users don't inadvertantly prevent older clients from interacting with the
cluster.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-25 14:53:14 -04:00
Sage Weil
eebd9a1bc2
osd/OSDMap: ignore pg_remap[_items] mappings when target osd isout
...
If the remap target is fully out, ignore the mapping.
Do not do a probabilistic 'osd_weight' rejection; pg-remap is intended to
avoid probabilistic mappings.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-25 14:53:14 -04:00
Sage Weil
27d6f4373b
osd/OSDMap: add pg_remap and pg_remap_items
...
The first simply forces a pg to map to a particular value. The second
replaces specific OSDs with specific other OSDs in the CRUSH mapping.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-25 14:53:14 -04:00
Sage Weil
e0b90d93c4
osd/OSDMap: clean up primary selection
...
_pg_to_raw_osds was selecting a primary, but it was generally
getting thrown out since _raw_to_up_osds was picking a *new*
primary after filtering out down and nonexistent osds. Add a
_pick_primary helper to do it and call that explicitly in the
callers where it can be used by all.
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-25 14:53:14 -04:00
Sage Weil
d8d24be869
Merge pull request #13965 from liewegas/wip-bluestore-pc
...
os/bluestore: fix perf counters
2017-03-25 13:08:33 -05:00
Sage Weil
42f3508a4b
Merge pull request #13962 from Liuchang0812/wip-add-override-in-osd-headers
...
osd: add override in headers files
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-25 13:07:37 -05:00
liuchang0812
acec741706
osd: add override in headers files
...
Signed-off-by: liuchang0812 <liuchang0812@gmail.com>
2017-03-25 12:55:38 +08:00
Kefu Chai
e32b1b8793
Merge pull request #14114 from dmick/wip-boost-j
...
debian/rules, ceph.spec.in: invoke cmake with -DBOOST_J
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-25 12:13:17 +08:00
Sage Weil
2d32138670
Merge pull request #13889 from liewegas/wip-denc-nullptr
...
include/denc: remove nullptr runtime magic boundedness check
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-24 16:41:37 -05:00
Sage Weil
3d3d2def66
Merge pull request #14096 from baiyanchun/remove_useless_parameter
...
common: remove useless parameter
Reviewed-by: Kefu Chai <kchai@redhat.com>
Reviewed-by: Pan Liu <liupan1111@gmail.com>
2017-03-24 16:41:18 -05:00
Sage Weil
50c9fd74f4
Merge pull request #14131 from liewegas/wip-crush-encode
...
crush: only encode class info if SERVER_LUMINOUS
Reviewed-by: Loic Dachary <ldachary@redhat.com>
2017-03-24 15:28:27 -05:00
Sage Weil
d44d859630
Merge pull request #13960 from wangzhengyong/kstore
...
os/kstore: some error handling
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-24 13:17:39 -05:00
Sage Weil
376a0b10aa
Merge pull request #13973 from shinobu-x/wp-sk-primarylogpg-null-nullptr
...
osd/PrimaryLogPG: nullptr not NULL
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-24 13:16:58 -05:00
Sage Weil
cb6efd842c
Merge pull request #13995 from liuhongtong/wip-config
...
common/config: set rocksdb_cache_size to OPT_U64
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-24 13:13:39 -05:00
Sage Weil
3283b284f8
Merge pull request #14013 from ShiqiCooperation/newshiqi
...
test/unittest_bluefs: check whether add_block_device success
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-24 13:12:16 -05:00
Sage Weil
cadd94cb3a
crush: only encode class info if SERVER_LUMINOUS
...
This fixes OSDMap reencode crc mismatches on jewel to
luminous upgrades.
Fixes: http://tracker.ceph.com/issues/19361
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-24 13:59:34 -04:00
Dan Mick
e7485da1fa
ceph.spec.in: derive _smp_ncpus and use it for -DBOOST_J
...
Signed-off-by: Dan Mick <dan.mick@redhat.com>
2017-03-24 10:53:52 -07:00
Dan Mick
1524daab38
ceph.spec.in: move lowmem_build setting of _smp_mflags
...
Signed-off-by: Dan Mick <dan.mick@redhat.com>
2017-03-24 10:53:52 -07:00
Dan Mick
6a9e81057b
debian/rules: invoke cmake with -DBOOST_J
...
Allow boost build during toplevel cmake from Debian package build
to benefit from multiple processors. Should speed build a lot
on many-proc machines (say, arm64). Use argument passed to
debhelper.
Signed-off-by: Dan Mick <dan.mick@redhat.com>
2017-03-24 10:53:24 -07:00
Casey Bodley
543ba8b835
Merge pull request #14082 from idealguo/update-bucket-acl
...
rgw: enable to update acl of bucket created in slave zonegroup
Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-03-24 11:15:05 -04:00
Casey Bodley
433295ccf6
Merge pull request #14043 from zhangsw/fix-rgw-deletebucket
...
rgw: delete non-empty buckets in slave zonegroup works not well
Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-03-24 11:11:50 -04:00
Casey Bodley
c899d7ef8c
Merge pull request #13991 from Liuchang0812/wip-rgw-optimization
...
rgw: avoid listing user buckets for rgw_delete_user
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-03-24 11:10:28 -04:00
Casey Bodley
bfa315f1c2
Merge pull request #13504 from rzarzynski/wip-rgw-chunkingfilter-cleanup
...
rgw: clean up the unneeded rgw::io::ChunkingFilter::has_content_length.
Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-03-24 11:08:18 -04:00
Kefu Chai
8fbf6d8f44
Merge pull request #13847 from wjwithagen/wip-wjw-ceph-disk-tests-2
...
ceph-disk/tests/test_main.py: FreeBSD does not do multipath
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-24 22:44:15 +08:00
Kefu Chai
21d8a97d19
Merge pull request #13974 from tchaikov/wip-vstart-start-mgr
...
vstart: do not start mgr if not start_all
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-24 21:44:56 +08:00
Kefu Chai
c52431b390
Merge pull request #13197 from asheplyakov/master-18740
...
systemd/ceph-disk: make it possible to customize timeout
Reviewed-by: Loic Dachary <ldachary@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-24 15:53:17 +08:00
Kefu Chai
6a771d1ef5
Merge pull request #14103 from tchaikov/wip-https-github
...
script: ceph-release-notes: use https instead of http
Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>
2017-03-24 14:43:48 +08:00
Sage Weil
bb8f633f5d
Merge pull request #14085 from wjwithagen/wip-wjw-bluestore-fixture
...
test/objectstore/store_test_fixture.cc: Exclude bluestore code if required.
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-23 20:47:45 -05:00
Sage Weil
117a63277c
Merge pull request #13931 from wangzhengyong/extent
...
os/bluestore: fix bug for calc extent_avg in reshard function
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
2017-03-23 20:47:12 -05:00
Sage Weil
8275fda6de
Merge pull request #14073 from liewegas/wip-bluestore-nullptr
...
os/bluestore: avoid nullptr in bluestore_extent_ref_map_t::bound_encode
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-23 20:44:59 -05:00
Sage Weil
c820749a04
Merge pull request #13577 from yonghengdexin735/wip-zzz-openalloc
...
os/bluestore: fix bug in _open_alloc()
Reviewed-by: Varada Kari <varada.kari@sandisk.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-23 20:44:35 -05:00
Loic Dachary
709dc377c7
Merge pull request #14110 from dachary/wip-crush-cleanup
...
crush: builder: clean the arguments of crush_reweight* methods
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Loic Dachary <ldachary@redhat.com>
2017-03-23 21:48:00 +01:00
Kefu Chai
b4f3a298f6
vstart.sh: do not init fsmap if "$new == 0"
...
we cannot create a new cephfs using a non-empty pool without '--force'
option now, so the "ceph fs new" command fails with "vstart.sh -k".
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-03-24 00:55:40 +08:00
Kefu Chai
0578177569
tests: remove mds,osd,mon args passed to vstart.sh
...
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-03-24 00:55:40 +08:00
Sahid Orentino Ferdjaoui
2c31903355
crush: builder: clean the arguments of crush_reweight* methods
...
This commit is just a cleanup to make the arguments of the method
around crush_reweight all coherent.
Signed-off-by: Sahid Orentino Ferdjaoui <sahid.ferdjaoui@redhat.com>
2017-03-23 17:54:52 +01:00
Kefu Chai
c9c2cf6b8a
vstart.sh: remove start_*
...
so there are only two ways to override the number of daemons to start
- using the env var CEPH_NUM_{MON|OSD|MGR|MDS} or {MON|OSD|MGR|MDS}
- command line options: --{mon,osd,mds}_num
do prevent a daemon from running, set the corrresponding env var to 0.
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-03-24 00:54:35 +08:00
Yuri Weinstein
18ca002958
Merge pull request #14050 from ovh/bp-dump-ops-by-duration
...
common/TrackedOp: allow dumping historic ops sorted by duration
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-23 08:47:55 -07:00
Yuri Weinstein
0620e81581
Merge pull request #14060 from LiumxNL/wip-170321
...
osd: combine unstable stats with info.stats when publish stats to osd
Reviewed-by: Sage Weil <sage@redhat.com>
2017-03-23 08:46:36 -07:00
Yuri Weinstein
c54eb54a2a
Merge pull request #13293 from Liuchang0812/cleanup-coverity
...
test, osd: fix some coverity issues
Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-03-23 08:45:58 -07:00
Casey Bodley
e7093c1e96
Merge pull request #14014 from Liuchang0812/wip-fix-seg-fault
...
rgw: fix memory leak in RGWGetObjLayout
Reviewed-by: Jos Collin <jcollin@redhat.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
2017-03-23 09:54:47 -04:00
Sage Weil
1ed62076a3
os/bluestore: avoid nullptr in bluestore_extent_ref_map_t::bound_encode
...
Signed-off-by: Sage Weil <sage@redhat.com>
2017-03-23 08:21:39 -05:00
Haomai Wang
93de4c60e6
Merge pull request #14094 from optimistyzy/322
...
bluestore, NVMeDevice: use task' own lock for (random) read
Reviewed-by: Haomai Wang <haomai@xsky.com>
2017-03-23 19:23:34 +08:00