[Ocfs2-tools-commits] branch, full-hb-ctl, created. ocfs2-tools-1.4.0-79-g7fe6833

[svn-commits at oss.oracle.com]

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, full-hb-ctl has been created
        at  7fe68335668f1ca468e35066747abe11152f7ea9 (commit)

- Log -----------------------------------------------------------------
commit 7fe68335668f1ca468e35066747abe11152f7ea9
Author: Joel Becker <joel.becker at oracle.com>
Date:   Tue Jun 3 14:56:09 2008 -0700

    ocfs2_hb_ctl: Enable support for all cluster stacks
    umount.ocfs2: Gone!
    
    Remove the check for non-o2cb cluster stacks in ocfs2_hb_ctl - it now
    is used for all stacks.
    
    With this, umount.ocfs2 is no longer needed.
    
    Signed-off-by: Joel Becker <joel.becker at oracle.com>

commit cc3c4e7887b513c0857f55e11801dbd3356a0eb7
Author: Joel Becker <joel.becker at oracle.com>
Date:   Tue Jun 3 15:08:30 2008 -0700

    mount.ocfs2,ocfs2_hb_ctl: Use OCFS2_FS_NAME as the mount service
    
    Use OCFS2_FS_NAME as the service for actual mounts.
    
    Signed-off-by: Joel Becker <joel.becker at oracle.com>

commit 9f0749ec08cd63aef218d51432de57f794b80305
Author: Joel Becker <joel.becker at oracle.com>
Date:   Tue Jun 3 16:16:17 2008 -0700

    ocfs2_controld: Allow multiple real mounts.
    
    We want all real mounts to advertise the same 'service', so the daemon
    must notice them and allow multiple mounts.  Real mounts will use the
    service OCFS2_FS_NAME.  Multiple instances of other services are still
    not allowed.
    
    Signed-off-by: Joel Becker <joel.becker at oracle.com>

commit a3c73b59fb8d7cc0efa836ed3721df940d33bbd2
Author: Joel Becker <joel.becker at oracle.com>
Date:   Tue Jun 3 15:54:31 2008 -0700

    ocfs2_controld: Rename 'mountpoint' to 'service'
    
    We really want to track services now, so rename 'struct mountpoint' to
    'struct service' in ocfs2_controld/mount.c
    
    Signed-off-by: Joel Becker <joel.becker at oracle.com>

commit d9a63a29e754d1b6fd63c2a07be669b23f9e828f
Author: Joel Becker <joel.becker at oracle.com>
Date:   Mon Jun 2 18:05:31 2008 -0700

    ocfs2_hb_ctl: Support all cluster stacks.
    
    ocfs2_hb_ctl now looks up the cluster descriptor, so it should be able
    to control group join/leave for all stacks.
    
    Signed-off-by: Joel Becker <joel.becker at oracle.com>

commit 332245be2d02ffb1c4ae19315141d7d37dab0f47
Author: Joel Becker <joel.becker at oracle.com>
Date:   Mon Jun 2 15:04:46 2008 -0700

    ocfs2_hb_ctl: Use ocfs2_hb_ctl again for o2cb umount(2).
    
    We've determined that mountpoints cannot be safely tracked from
    userspace.  The only safe place to leave a filesystem group is in
    put_super(7).  We used to do this via ocfs2_hb_ctl, and we're going to
    do so again.
    
    This change does three things:
    
    1) Puts ocfs2_hb_ctl(8) back in /proc/sys/fs/ocfs2/nm/hb_ctl_path
    2) Teaches ocfs2_hb_ctl(8) to do nothing when the cluster stack isn't o2cb
    3) Teaches umount.ocfs2(8) to skip cluster interaction when the cluster
       stack *is* o2cb.
    
    Thus, ocfs2_hb_ctl(8) is used for o2cb clusters and umount.ocfs2(8) is
    used for all other cluster stacks.  Future changes will add support for
    all cluster stacks to ocfs2_hb_ctk(8), removing the need for
    umount.ocfs2(8).
    
    Signed-off-by: Joel Becker <joel.becker at oracle.com>

-----------------------------------------------------------------------


hooks/post-receive
-- 
Tools to manage the ocfs2 filesystem.