15a20eb...
by
Sage Weil
on 2017-10-05
Merge pull request #18120 from liewegas/ wip-bluefs- allocation
os/bluestore: make bluefs behave better near enospc
Reviewed-by: xie xingguo <email address hidden>
667e14e...
by
Mikolaj Golub
on 2017-10-05
Merge pull request #18114 from dillaman/wip-21670
librbd: list_children should not attempt to refresh image
Reviewed-by: Mykola Golub <email address hidden>
7d1c99f...
by
Sage Weil
on 2017-10-05
Merge pull request #18118 from liewegas/wip-21470
os/bluestore: fix another aio stall/deadlock
Reviewed-by: Varada Kari <email address hidden>
Reviewed-by: xie xingguo <email address hidden>
f9cfe84...
by
Sage Weil
on 2017-10-05
Merge pull request #18105 from gregsfortytwo/ wip-sigpipe- porting
msg: reimplement sigpipe blocking
Reviewed-by: Sage Weil <email address hidden>
4341f27...
by
Sage Weil
on 2017-10-05
Merge pull request #17710 from Songweibin/ wip-add- pybind
pybind/rados: add support open_ioctx2 API
Reviewed-by: Josh Durgin <email address hidden>
f26c3c7...
by
Sage Weil
on 2017-10-05
Merge pull request #18079 from liewegas/wip-21618
qa/standalone/ scrub/osd- scrub-repair: no -y to diff
Reviewed-by: David Zafman <email address hidden>
75dc9b6...
by
Sage Weil
on 2017-10-05
Merge pull request #18088 from dillaman/wip-21629
osd: additional protection for out-of-bounds EC reads
Reviewed-by: Sage Weil <email address hidden>
65be614...
by
Sage Weil <email address hidden>
on 2017-10-03
os/bluestore: add bluestore_ bluefs_ min_free
We need at least ~1GB free so we can write out new SSTs (which are 256MB
each).
Signed-off-by: Sage Weil <email address hidden>
569e924...
by
Sage Weil <email address hidden>
on 2017-10-03
os/bluestore/ BlueFS: crash on enospc
We were passing this error back to rocksdb, which would then crash (much)
later with a corrupted SST file! No good.
Signed-off-by: Sage Weil <email address hidden>
67ec758...
by
Sage Weil <email address hidden>
on 2017-10-03
os/bluestore: use normal Context for async deferred_try_submit
I'm not quite sure why the FunctionContext did not ever execute on the
finisher thread (perhaps the [&] captured some state on the stack that it
shouldn't have?). In any case, using a traditional Context here appears
to resolve the problem (of the async deferred_ try_submit( ) never executing,
leading to a bluestore stall/deadlock).
Fixes: http:// tracker. ceph.com/ issues/ 21470
Signed-off-by: Sage Weil <email address hidden>