github/containerd
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 2 Wiki Activity

Fix race with task checkpoint

Browse Source 
Because runc will delete a container after a successful checkpoint we
need to handle a NotFound error from runc on delete.

There is also a race between SIGKILL'ing the shim and it actually
exiting to unmount the tasks rootfs, we need to loop and wait for the
task to actually be reaped before trying to delete the rootfs+bundle.

Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
This commit is contained in:
Michael Crosby
2017-08-09 15:44:42 -04:00
parent 258c719cbc
commit d513dd2bfd
5 changed files with 43 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
linux/shim/init.go
View File

@@ -237,31 +237,33 @@ func (p *initProcess) SetExited(status int) {
}
func (p *initProcess) Delete(context context.Context) error {
status, err := p.Status(context)
if err != nil {
return err
}
if status != "stopped" {
return fmt.Errorf("cannot delete a running container")
}
p.killAll(context)
p.Wait()
err = p.runtime.Delete(context, p.id, nil)
err := p.runtime.Delete(context, p.id, nil)
// ignore errors if a runtime has already deleted the process
// but we still hold metadata and pipes
//
// this is common during a checkpoint, runc will delete the container state
// after a checkpoint and the container will no longer exist within runc
if err != nil {
if strings.Contains(err.Error(), "does not exist") {
err = nil
} else {
err = p.runtimeError(err, "failed to delete task")
}
}
if p.io != nil {
for _, c := range p.closers {
c.Close()
}
p.io.Close()
}
err = p.runtimeError(err, "OCI runtime delete failed")
if err2 := mount.UnmountAll(p.rootfs, 0); err2 != nil {
log.G(context).WithError(err2).Warn("Failed to cleanup rootfs mount")
log.G(context).WithError(err2).Warn("failed to cleanup rootfs mount")
if err == nil {
err = errors.Wrap(err2, "Failed rootfs umount")
err = errors.Wrap(err2, "failed rootfs umount")
}
}
return err
}