Messages in this thread | | | Date | Tue, 29 Jul 2014 11:39:07 -0400 | From | Rik van Riel <> | Subject | Re: [PATCH 1/2] sched: fix and clean up calculate_imbalance |
| |
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 07/29/2014 11:31 AM, Vincent Guittot wrote: > On 29 July 2014 16:53, Rik van Riel <riel@redhat.com> wrote: >> I cannot think of any case where keeping the "overloaded" code >> would result in the code behaving differently over the long >> term. > >> What am I overlooking? > > IIUC the load_above_capacity is there to prevent the busiest group > to become idle and you remove that protection
Having the busiest group become partially idle is perfectly fine, as long as there is also spare capacity in the destination group.
Spreading out the work load over more CPU sockets, with more CPU cache and more total available memory bandwidth is often a good thing.
- -- All rights reversed -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iQEcBAEBAgAGBQJT18AbAAoJEM553pKExN6DaLoH/RzCR/YWDYe4/tZEC/sHn89l jCOVdoY2ZkVHx/yO1Vd0UeoAE94XFsiw0sKwy0lVXLxi3nBIjfh41qZkGoS2lQuo YSi7PE2o6T5WmluEQfTnUn1DAX0sigUYvWAxnpDYM6I+m8RS0QB7VYSaoqaT3T8D qomO8FmBN75JrwoywWBDiVRjnxfQJzt6bBK92GUIo6+af6RUSQ9xwQg6EPp3E3r6 FFqIHJ6HaRFCml1KMD4azopVC/bbbhvnbE/EO/5In2nbBsVK06XpBlLWfCyiD48q NkjnAjISweWTCoMdhULaXdLGeWH85DWoDtvDTCbK+m/QBXRXdiAbij0fDFDhPck= =QIq6 -----END PGP SIGNATURE-----
| |