← Back to team overview

yahoo-eng-team team mailing list archive

[Bug 2109675] [NEW] With CPU pinning enabled and the flavor’s NUMA node set to 1 by default, scheduling fails when the NUMA node becomes fully utilized.

 

Public bug reported:

1.numa node:
available: 2 nodes (0-1)
node 0 cpus: 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38
node 0 size: 96058 MB
node 0 free: 62410 MB
node 1 cpus: 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
node 1 size: 96736 MB
node 1 free: 58459 MB

2.When NUMA node 0 is fully utilized, virtual machines are not scheduled
to NUMA node 1.

3.ERROR:UnboundLocalError: "local variable 'sibling_set' referenced
before assignment"

4.Built sibling_sets: defaultdict(<type 'list'>, {}), when sibling_sets
is empty, should return

** Affects: nova
     Importance: Undecided
     Assignee: LiuWenYu (liuwenyu)
         Status: New

** Changed in: nova
     Assignee: (unassigned) => LiuWenYu (liuwenyu)

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Compute (nova).
https://bugs.launchpad.net/bugs/2109675

Title:
  With CPU pinning enabled and the flavor’s NUMA node set to 1 by
  default, scheduling fails when the NUMA node becomes fully utilized.

Status in OpenStack Compute (nova):
  New

Bug description:
  1.numa node:
  available: 2 nodes (0-1)
  node 0 cpus: 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38
  node 0 size: 96058 MB
  node 0 free: 62410 MB
  node 1 cpus: 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
  node 1 size: 96736 MB
  node 1 free: 58459 MB

  2.When NUMA node 0 is fully utilized, virtual machines are not
  scheduled to NUMA node 1.

  3.ERROR:UnboundLocalError: "local variable 'sibling_set' referenced
  before assignment"

  4.Built sibling_sets: defaultdict(<type 'list'>, {}), when
  sibling_sets is empty, should return

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/2109675/+subscriptions