On Mon, Sep 25, 2017 at 02:53:01PM +0100, Daniel P. Berrange wrote:
On Thu, Sep 14, 2017 at 02:41:12AM +0200, Wojtek Porczyk wrote:
> The intended use is to ensure that the implementation is empty, which is
> one way to ensure that all connections were properly closed and file
> descriptors reclaimed.
>
> Signed-off-by: Wojtek Porczyk <woju(a)invisiblethingslab.com>
> ---
(snip)
> + @asyncio.coroutine
> + def drain(self):
> + '''Wait for the implementation to become idle.
> +
> + This is a coroutine.
> + '''
> + self.log.debug('drain()')
> + if self._pending:
> + yield from self._finished.wait()
> + self.log.debug('drain ended')
What is responsible for calling 'drain' ?
Users of the library, and they do it at their pleasure. This is to allow the
loop to actually run the scheduled tasks. After calling virConnect.close() the
handles/timeouts aren't actually closed until the loop run the scheduled
callbacks. In practice a simple `yield` in a coroutine would be also
sufficient since respective Tasks are in the _ready queue and all run during
next loop cycle, but that's not guaranteed in asyncio specification.
--
pozdrawiam / best regards _.-._
Wojtek Porczyk .-^' '^-.
Invisible Things Lab |'-.-^-.-'|
| | | |
I do not fear computers, | '-.-' |
I fear lack of them. '-._ : ,-'
-- Isaac Asimov `^-^-_>