On 12/21/22 20:36, Kevin Traynor wrote:
> Hi Cheng,
>
> On 17/12/2022 13:15, Cheng Li wrote:
>> Currently, pmd_rebalance_dry_run() calculate overall variance of
>> all pmds regardless of their numa location. The overall result may
>> hide un-balance in an individual numa.
>>
>> Considering the following case. Numa0 is free because VMs on numa0
>> are not sending pkts, while numa1 is busy. Within numa1, pmds
>> workloads are not balanced. Obviously, moving 500 kpps workloads from
>> pmd 126 to pmd 62 will make numa1 much more balance. For numa1
>> the variance improment will be almost 100%, because after rebalance
>
> typo: improvement - maybe maintainer can fix on apply and punctuate> commit
> title ("dpif-netdev: Calculate per numa variance.") :-)
I have a git hook that capitalizes the first letter and adds a period
at the end for every patch that I'm applying. :)
>
>> each pmd in numa1 holds same workload(variance ~= 0). But the overall
>> variance improvement is only about 20%, which may not tigger auto_lb.
>>
>> ```
>> numa_id core_id kpps
>> 0 30 0
>> 0 31 0
>> 0 94 0
>> 0 95 0
>> 1 126 1500
>> 1 127 1000
>> 1 63 1000
>> 1 62 500
>> ```
>>
>> As auto_lb doesn't balance workload across numa nodes. So it makes
>> more sense to calculate variance improvemnet per numa node.
>>
>> Signed-off-by: Cheng Li <[email protected]>
>> Signed-off-by: Kevin Traynor <[email protected]>
>> Co-authored-by: Kevin Traynor <[email protected]>
>
>
> I think it's a nice improvement, thanks for sending it. I reviewed and tested
> a bit and the debug logs look good now [1].
>
> Acked-by: Kevin Traynor <[email protected]>
Applied. Thanks!
Best regards, Ilya Maximets.
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
