Prior to this change, there were double-colon ("::") characters in the
HTML docs to indicate literal blocks.
The proceeding paragraphs were not pre-formatted because there was no
newline between the double-colon and the to-be-pre-formatted paragraph.
Move the "::" characters inline to simplify the syntax and make Sphinx
properly interpret these texts as literal blocks.
Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
Introduces Mantle, a programmable metadata load balancer. Policies for making
migration decisions are written in Lua but the Migrator and Balancer modules
still do fragmentation and migration. If the Lua balancer fails, control falls
back to the original balancer implementation.
Signed-off-by: Michael Sevilla <mikesevilla3@gmail.com>
Without the patch front-ends issueed a lot of small IO operations
leading to increased overhead on syscalls and to the fragmentation
of an HTTP message across multiple TCP segments. The later was
occuring when the Nagle's algorithm hadn't been able to form
a single TCP segment (usually when running on extremely fast
network interfaces like loopback).
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
std::string and thus boost::string_ref ARE OBLIGED to carry multiple
0x00 and count them to the length of a string. We need to take that
into consideration and sanitize the size of a ceph::buffer::list used
to store metadata values (x-amz-meta-*, X-Container-Meta-*, etags).
Otherwise we might send 0x00 to clients.
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
Fix#17562 (backtrace check fails when scrubbing directory created by fsstress)
Reviewed-by: John Spray <john.spray@redhat.com>
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
Normally we never call encode on a message that has a byte_throttler set
because we only use it for messages we received. However, for forwarded
messages that we clear_payload() before resending, we *do* reencode, and in
that case we need to retake the appropriate number of bytes from the
throttler--just like we release them in clear_payload().
Signed-off-by: Sage Weil <sage@redhat.com>
This ensures we reencode the payload with the
appropriate set of features if the client, us, or the
target do not have identical features. Otherwise we
may forward an encoding with more features than the
target can handle.
Signed-off-by: Sage Weil <sage@redhat.com>
Attempting to retrieve the group spec will fail on older OSDs, so it
must be executed as an individual step in the refresh state machine.
Also fixed code style issues for out parameters.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Because of a missing return, ceph-disk prepare would fail if given a
regular file as a journal. If the journal file does not exist, ceph-disk
will create it but fail to ensure that the ceph user owns it. The
symlink to the journal file is not set when the journal file is
specified on the command line and the journal file does not exist at
all. The ceph-osd daemon will silently create it as a file but it will
not be the file given in argument.
Add a test case to verify using a regular file as a journal works as
expected.
Fixes: http://tracker.ceph.com/issues/17662
Signed-off-by: Jayashree Candadai <jayaajay@indiana.edu>
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Two patches both added these definitions
in slightly different places, so git let them
merge without a conflict.
Signed-off-by: John Spray <john.spray@redhat.com>
