[PATCH sandbox 0/4] Fix data retention with QEMU sandboxes

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



In testing the docker sandbox patches I found that with QEMU
sandboxes, we would randomly loose data written to the block
device. This meant that after untar'ing the image contents into
the qcow2 the filesystem was left empty!

Eventually I realized that there was no where we are calling sync()
to flush data out to disk. We never noticed this before because
the sandbox code mostly used 9p filesystems with QEMU which do not
need this sync().

Daniel P. Berrange (4):
  push changing of user ID down into child process
  Sync and unmount filesystems during shutdown
  Fix passing of strace option to guest kernel
  Don't close immediately when getting EOF on RPC console

 bin/virt-sandbox.c                                |  23 ++-
 libvirt-sandbox/libvirt-sandbox-builder-machine.c |   7 +-
 libvirt-sandbox/libvirt-sandbox-init-common.c     | 226 +++++++++++++++++++---
 libvirt-sandbox/libvirt-sandbox-init-qemu.c       |   3 +-
 4 files changed, 225 insertions(+), 34 deletions(-)

-- 
2.4.3

--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list



[Index of Archives]     [Virt Tools]     [Libvirt Users]     [Lib OS Info]     [Fedora Users]     [Fedora Desktop]     [Fedora SELinux]     [Big List of Linux Books]     [Yosemite News]     [KDE Users]     [Fedora Tools]