At Wed, 30 Jul 2014 16:24:35 +0900, Hitoshi Mitake wrote: > > This patch revives VDI locking mechanism. When two or more clients > (QEMU and tgt) try to open one VDI, sheep returns an error to the > later one. > > Example: > $ sudo qemu-system-x86_64 -hda sheepdog:debian > qemu-system-x86_64: -hda sheepdog:debian: could not open disk image > sheepdog:debian: cannot get vdi info, VDI isn't locked, debian 0 > > v2: > - Changes for QEMU side are needless > - Manual unlocking is required for keeping data integrity > -- New dog subcommands are provided for this purpose > > Hitoshi Mitake (7): > sheep: change a prototype of process_main() for obtaining sender > information > sheep: revive lock operation > dog: use GET_VDI_INFO unconditionally in dog > sheep: snapshot and collect vdi state during joining to cluster > sheep: a new helper function for formatting struct node_id > dog: add a new subcommand "dog vdi lock list" for listing vdi locking > status > dog: add a new subcommand "dog vdi lock force-unlock" for unlocking > VDI forcibly > > dog/treeview.c | 41 ++++++++ > dog/treeview.h | 2 + > dog/vdi.c | 96 +++++++++++++++++- > include/internal_proto.h | 23 +++++ > include/sheep.h | 11 +- > include/sheepdog_proto.h | 6 ++ > sheep/group.c | 165 ++++++++++++++++++++++++++++-- > sheep/ops.c | 177 +++++++++++++++++++++++++-------- > sheep/request.c | 3 +- > sheep/sheep.c | 2 + > sheep/sheep_priv.h | 24 +++-- > sheep/vdi.c | 254 > +++++++++++++++++++++++++++++++++++++++++++++++ > 12 files changed, 736 insertions(+), 68 deletions(-)
Valerio, Fabian, could you test this series? You can pull the same thing from our repository: https://github.com/sheepdog/sheepdog/tree/vdi-locking-v2 In addition, you can use existing QEMU. Thanks, Hitoshi -- sheepdog mailing list sheepdog@lists.wpkg.org http://lists.wpkg.org/mailman/listinfo/sheepdog
