Note: This is a beta release of Red Hat Bugzilla 5.0. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Also email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback here.
Bug 1515787 - one core from each numa node should be reserved for the Host
Summary: one core from each numa node should be reserved for the Host
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: Red Hat OpenStack
Classification: Red Hat
Component: documentation
Version: 10.0 (Newton)
Hardware: Unspecified
OS: Unspecified
medium
medium
Target Milestone: async
: 10.0 (Newton)
Assignee: Sandra McCann
QA Contact: RHOS Documentation Team
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2017-11-21 11:59 UTC by David Juran
Modified: 2018-02-07 19:29 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2018-02-07 19:29:29 UTC


Attachments (Terms of Use)

Description David Juran 2017-11-21 11:59:48 UTC
Description of problem:
In the diagram in https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/10/html/network_functions_virtualization_planning_guide/ch_sriov-planning#concept_sriov-cpu-partitioning we describe that two hosts _from the same_ NUMA node are reserved for the host. In my opinion, it should be one core from each node. Because:

*) If you allocate two cores from NUMA 0, there will be an imbalance between the cores dedicated for VNF:s between the nodes. This could make it hard to schedule VNF:s, resulting in wasted cores

*) If I'm not mistaken, IRQ:s are routed to the first core in each NUMA node. A VNF pinned there will receive lots of interrupts.

Also, the section talks about Open vSwitch which I find confusing, considering this is about SR-IOV. I'm guessing this could be dropped?


Note You need to log in before you can comment on or make changes to this bug.