yahoo-eng-team team mailing list archive

[Brin Zhang <1874664@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

When boot more than one instance with accelerator, and the accelerators
are in one compute node, there will be two problems as below:

One problem is as we always get the first item(alloc_reqs[0]) in
alloc_reqs, when we iterator the second instance, it will throw conflict
exception when putting the allocations.

Another is as we always get the first item in
alloc_reqs_by_rp_uuid.get(selected_host.uuid), the selected_alloc_req is
always stable, that will cause the values in selections_to_return are
same . In fact, it's not right for subsequent operations.

More details you can see:
https://etherpad.opendev.org/p/filter_scheduler_issue_with_accelerators

** Affects: nova
     Importance: Medium
     Assignee: Wenping Song (wenping1)
         Status: Confirmed


** Tags: cyborg scheduler

** Changed in: nova
   Importance: Undecided => Medium

** Changed in: nova
       Status: New => Confirmed

** Changed in: nova
     Assignee: (unassigned) => Wenping Song (wenping1)

** Tags added: schedul

** Tags removed: schedul
** Tags added: cyborg scheduler

-- 
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/1874664

Title:
  Boot more than one instances failed with accelerators in its flavor

Status in OpenStack Compute (nova):
  Confirmed

Bug description:
  When boot more than one instance with accelerator, and the
  accelerators are in one compute node, there will be two problems as
  below:

  One problem is as we always get the first item(alloc_reqs[0]) in
  alloc_reqs, when we iterator the second instance, it will throw
  conflict exception when putting the allocations.

  Another is as we always get the first item in
  alloc_reqs_by_rp_uuid.get(selected_host.uuid), the selected_alloc_req
  is always stable, that will cause the values in selections_to_return
  are same . In fact, it's not right for subsequent operations.

  More details you can see:
  https://etherpad.opendev.org/p/filter_scheduler_issue_with_accelerators

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1874664/+subscriptions