[Ocfs2-tools-commits] branch, master, updated. ocfs2-tools-1.4.0-131-g2d9bac7
svn-commits at oss.oracle.com
svn-commits at oss.oracle.com
Thu Aug 14 15:53:59 PDT 2008
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "Tools to manage the ocfs2 filesystem.".
The branch, master has been updated
via 2d9bac7d21f98f9f2e6ef2de414c2f77e57a9d25 (commit)
via a538a7280e81fa7a5b0e06b6fc62f4867fccf4cd (commit)
via 57da442c8cf8cd87b85d795dccf927aff8cd30e5 (commit)
via 12d473254603f172231b60f1e602d01b372afe5c (commit)
via 13b17050fbd9fe007f55e2f9dd5603749815ca70 (commit)
via 9deaebbdbcb2c56f0eab60910348b6aae54aa6d7 (commit)
via 081552f4eadfbc7f07b7f326553e4c80440bc12a (commit)
via c58eaead0562d55d7ede24be14c29787e4c829a5 (commit)
via e9d7abb514e89a54a3756062912a953a97a891fa (commit)
via 4c10fd26bfb8ae9787b659fe233f2854f940ee66 (commit)
from 8658721b21ab58dcac63f144bbc30ed4832cb540 (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
- Log -----------------------------------------------------------------
commit 2d9bac7d21f98f9f2e6ef2de414c2f77e57a9d25
Author: Joel Becker <joel.becker at oracle.com>
Date: Tue Aug 12 15:59:05 2008 -0700
ocfs2_controld: Notify dlm_controld when we have a node go down.
ocfs2 must start recovery before the dlm completes its recovery.
Otherwise, the dlm might give out locks that belonged to the dead node
before ocfs2 has had a chance to replay its journal.
libdlmcontrol implements this ordering. When ocfs2_controld sees a node
go down, it tells dlm_controld via dlmc_fs_notified(). dlm_controld
responds with success if it also has seen the node go down. It
responds nonzero otherwise. dlm_controld knows to wait until it has
seen the notification from ocfs2_controld.
The ocfs2_controld implementation is simple. It first notifies ocfs2,
then sends the notification to dlm_controld. ocfs2_controld will keep
resending it while dlm_controld responds in the negative. When
dlm_controld finally gets the node down event, ocfs2_controld will stop
sending. We keep such a simple busy-looping scheme because the
back-and-forth traffic actually means context switches and process
sleeping.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit a538a7280e81fa7a5b0e06b6fc62f4867fccf4cd
Author: Joel Becker <joel.becker at oracle.com>
Date: Tue Aug 12 15:09:15 2008 -0700
ocfs2_controld: De-register all filesystems on exit.
If we're exiting uncleanly, we'll have a list of registered filesystems.
Send fire-and-forget unregister events to dlm_controld.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit 57da442c8cf8cd87b85d795dccf927aff8cd30e5
Author: Joel Becker <joel.becker at oracle.com>
Date: Fri Aug 1 19:55:23 2008 -0700
ocfs2_controld: Register mountgroups with dlm_controld
Links up the mountgroup code with the dlmcontrol code.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit 12d473254603f172231b60f1e602d01b372afe5c
Author: Joel Becker <joel.becker at oracle.com>
Date: Fri Aug 1 19:13:29 2008 -0700
ocfs2_controld: Registration function for lockspaces.
We want to register a filesystem uuid with dlm_controld so that the
lockspace waits on us before recovery. We use libdlmcontrol's
registration functions.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit 13b17050fbd9fe007f55e2f9dd5603749815ca70
Author: Joel Becker <joel.becker at oracle.com>
Date: Fri Aug 1 19:13:03 2008 -0700
ocfs2_control: udev rule update
We only have one filename, so there is no need to use %k. udevd didn't
like it somehow.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit 9deaebbdbcb2c56f0eab60910348b6aae54aa6d7
Author: Joel Becker <joel.becker at oracle.com>
Date: Fri Aug 1 15:52:26 2008 -0700
ocfs2_controld: Connect to dlm_controld.
Our first step in introducing the dlmcontrol API is to connect and
disconnect from dlm_controld at start and exit.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit 081552f4eadfbc7f07b7f326553e4c80440bc12a
Author: Joel Becker <joel.becker at oracle.com>
Date: Tue Jul 29 18:15:01 2008 -0700
ocfs2_controld: Make libdlmcontrol required
ocfs2_controld needs to use libdlmcontrol to talk to dlm_controld. This
makes configure detect its presence. It also separates
BUILD_OCFS2_CONTROLD from BUILD_CMAN_SUPPORT. The parts required for
all ocfs2_controlds (cpg, openais, libdlm, libdlmcontrol) are tested
separately then the cman part specific to a cman controld.
This allows other stacks to check and set BUILD_XXX_SUPPORT for their
own controld independent of cpg/libdlm.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit c58eaead0562d55d7ede24be14c29787e4c829a5
Author: Joel Becker <joel.becker at oracle.com>
Date: Fri Aug 8 18:07:50 2008 -0700
libo2dlm: Fix a couple of bugs accessing libdlm.
This fixes two bugs with libdlm. First, older versions of fs/dlm could
return EEXIST while the filesystem is mounted. Handle that. Second,
and uninitialized dlm_ctxt sometimes prevented us from loading libdlm in
the first place.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit e9d7abb514e89a54a3756062912a953a97a891fa
Author: Joel Becker <joel.becker at oracle.com>
Date: Tue Jul 29 17:08:36 2008 -0700
libo2dlm: We create libdlm.h for compatibility when we don't find libdlm. Let's ignore that file.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
commit 4c10fd26bfb8ae9787b659fe233f2854f940ee66
Author: Joel Becker <joel.becker at oracle.com>
Date: Thu Aug 14 12:14:37 2008 -0700
ocfs2_controld,debugfs.ocfs2: Dump the daemon debug buffer via debugfs.ocfs2.
ocfs2_controld maintains a circular buffer of debug messages. Until
now, the only way to see it was to run in the foreground via the -D
switch. That's annoying to have to set up, and not useful in a
production environment.
We introduce the DUMP message to the controld protocol. This message
requests the entire debug buffer. ocfs2_controld sends the buffer as a
LIST reply. Each ITEM in the list is one OCFS2_CONTROLD_MAXLINE packet.
Thus, the ITEMCOUNT is (roughly) "size of dump buffer /
OCFS2_CONTROLD_MAXLINE".
libo2cb gains the o2cb_control_daemon_debug() API. This function sends
the DUMP message, receives the list, and returns a single string with
the entire buffer concatenated back together.
debugfs.ocfs2 gains the "controld" command. Currently it only has one
sub-operation, "dump". This calls o2cb_control_daemon_debug(), then
prints the results via the pager.
Signed-off-by: Joel Becker <joel.becker at oracle.com>
Signed-off-by: Mark Fasheh <mfasheh at suse.com>
-----------------------------------------------------------------------
Summary of changes:
Config.make.in | 1 +
configure.in | 33 +++-
debugfs.ocfs2/commands.c | 44 +++++-
debugfs.ocfs2/main.c | 5 +-
include/o2cb/o2cb.h | 2 +
include/o2cb/o2cb_client_proto.h | 1 +
libo2cb/client_proto.c | 1 +
libo2cb/o2cb_abi.c | 77 ++++++++
libo2dlm/.gitignore | 1 +
libo2dlm/o2dlm.c | 13 +-
libo2dlm/o2dlm_err.et | 3 +
ocfs2_controld/Makefile | 10 +-
ocfs2_controld/dlmcontrol.c | 366 ++++++++++++++++++++++++++++++++++++++
ocfs2_controld/main.c | 144 +++++++++------
ocfs2_controld/mount.c | 66 +++++++-
ocfs2_controld/ocfs2_controld.h | 9 +
vendor/common/51-ocfs2.rules | 2 +-
17 files changed, 696 insertions(+), 82 deletions(-)
create mode 100644 ocfs2_controld/dlmcontrol.c
hooks/post-receive
--
Tools to manage the ocfs2 filesystem.
More information about the Ocfs2-tools-commits
mailing list