Displaying 3 results from an estimated 3 matches for "num_in_flight".
2019 Aug 26
2
Re: [PATCH disk-sync 5/5] Convert disk_sync inner loop to asynchronous.
...ed.
>+ count += length
>+
>+ process_aio_requests(nbd_handle)
In order to allow less requests in flight, would it be enough to just do
something like this here (similarly to wait_for_aio_commands_to_finish)?
while nbd_handle.aio_in_flight() > NUM_IN_FLIGHT:
nbd_handle.poll(-1)
Also, I presume all of the locking is left to libnbd to be done (and as you can
see I don't concern myself with any locking in the whole file), but if that was
to be improved, is there some python part that would require it? For example
when cleaning up the code?...
2019 Aug 27
0
Re: [PATCH disk-sync 5/5] Convert disk_sync inner loop to asynchronous.
...t += length
> >+
> >+ process_aio_requests(nbd_handle)
>
> In order to allow less requests in flight, would it be enough to just do
> something like this here (similarly to wait_for_aio_commands_to_finish)?
>
> while nbd_handle.aio_in_flight() > NUM_IN_FLIGHT:
> nbd_handle.poll(-1)
Yes we could do this.
The eventual solution may be to replace the whole loop with a polling
loop, but this would be a more invasive change. For comparison of
this approach see:
https://github.com/libguestfs/libnbd/blob/7ef893735937cd7ae62d0b41171ec14195ef2710/ex...
2019 Aug 22
7
[PATCH disk-sync 0/5] Misc cleanups and convert inner loop to asynch.
This is based on top of:
https://github.com/nertpinx/v2v-conversion-host/commit/0bb2efdcacd975a2cae7380080991ac7fc238d2b
The first 4 patches are fairly uncontroversial miscellaneous cleanups.
Patch 5 is the interesting one. (Note it doesn't quite work yet, so
it's for discussion only.)
Patch 5 converts the inner loop to use asynchronous libnbd calls.
performance improves quite a bit for