TestServerNumaPCI may give false positives

Bug #1573282 reported by Waldemar Znoinski on 2016-04-21
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
intel-nfv-ci-tests
Medium
Waldemar Znoinski

Bug Description

NUMA and PCI Topology in host VM:
tempest@ubuntu:/opt/stack/new/tempest$ lstopo
Machine (16GB)
  NUMANode L#0 (P#0 7985MB) + Socket L#0
    L2 L#0 (4096KB) + Core L#0
      L1d L#0 (32KB) + L1i L#0 (32KB) + PU L#0 (P#0)
      L1d L#1 (32KB) + L1i L#1 (32KB) + PU L#1 (P#1)
    L2 L#1 (4096KB) + Core L#1
      L1d L#2 (32KB) + L1i L#2 (32KB) + PU L#2 (P#2)
      L1d L#3 (32KB) + L1i L#3 (32KB) + PU L#3 (P#3)
  NUMANode L#1 (P#1 8063MB) + Socket L#1
    L2 L#2 (4096KB) + Core L#2
      L1d L#4 (32KB) + L1i L#4 (32KB) + PU L#4 (P#4)
      L1d L#5 (32KB) + L1i L#5 (32KB) + PU L#5 (P#5)
    L2 L#3 (4096KB) + Core L#3
      L1d L#6 (32KB) + L1i L#6 (32KB) + PU L#6 (P#6)
      L1d L#7 (32KB) + L1i L#7 (32KB) + PU L#7 (P#7)
  HostBridge L#0
    PCI 8086:7010
    PCI 1013:00b8
    PCI 1af4:1000
    PCI 1af4:1001

TestServerNumaPCI schedules a guest VM + pci device in passthru mode. It compares whether the same socket is used for both CPUs, memory of spawned guest VM and I/O device that's passed thru. Given only one PCI HostBridge when VMs CPU and memory is on NUMA node 0 the test will succeed. This may give false positives if the VMs is always scheduled to NUMA node 0.

Waldemar Znoinski (wznoinsk) wrote :
Changed in intel-nfv-ci-tests:
assignee: nobody → Waldemar Znoinski (wznoinsk)
importance: Undecided → Medium
status: New → In Progress

Reviewed: https://review.openstack.org/309210
Committed: https://git.openstack.org/cgit/openstack/intel-nfv-ci-tests/commit/?id=153db3fdab2b299bf3bcf5ae315ad4653ea885a7
Submitter: Jenkins
Branch: master

commit 153db3fdab2b299bf3bcf5ae315ad4653ea885a7
Author: Waldemar Znoinski <email address hidden>
Date: Thu Apr 21 23:43:19 2016 +0100

    remove TestServerNumaPCI

    This change removes TestServerNumaPCI due to fact
    that there's only one PCI HostBridge in a host VM
    the test may be giving false positives if the guest VM
    is scheduled to NUMA Node 0.

    Change-Id: Ie3f87620841e024fd42eea4187e31d46c57e7e84
    Closes-Bug: #1573282

Changed in intel-nfv-ci-tests:
status: In Progress → Fix Released
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers