On 26.03.2025 15:30, Ilya Maximets wrote:
> On 3/25/25 00:21, Vladislav Odintsov wrote:
>> With this commit the rebalance of hash entries between bonding members
>> becomes less frequent.  If we know all bond members' speed, we do not
>> move hash entries between them if load difference is less than 1.5% of
>> minimum link speed of bond members.
>>
>> Reported-at: 
>> https://mail.openvswitch.org/pipermail/ovs-dev/2025-March/422028.html
>> Suggested-by: Ilya Maximets <i.maxim...@ovn.org>
>> Signed-off-by: Vladislav Odintsov <vlodintsov@k2.cloud>
>> ---
>> v1 -> v2:
>>    - Addressed Ilya's, Eelco's, Mike's review comments.
>>    - Docs updated.
>>    - NEWS entry added.
>> ---
>>   Documentation/topics/bonding.rst |  3 ++-
>>   NEWS                             |  4 ++++
>>   ofproto/bond.c                   | 25 +++++++++++++++++++------
>>   3 files changed, 25 insertions(+), 7 deletions(-)
> Hi, Vladislav.  Thanks for v2!
>
> Have you seen my reply for your observation about comparing speed to raw
> byte counters here:
>    https://mail.openvswitch.org/pipermail/ovs-dev/2025-March/422255.html
>
> Tl;DR; I think you're right and we need to change the way we count the load
> in order to be able to compare the load to a link speed.
>
> Best regards, Ilya Maximets.

Hi Ilya,

Thanks for review!

I changed the condition before calling a bond_shift_load() function, so 
now Bps with Bps instead of B with Bps should be compared.  It seems to 
me that this is good enough.

Do I understand you correctly, that you proposed to change the movement 
logic inside bond_shift_load() function?  If yes, I'm wondering why do 
we care, whether we compare total carried bytes during last rebalance 
interval or an average bytes per second during the same interval?  In 
both cases we seem to fairly compare same units: B vs B or Bps vs Bps, 
so the current logic of this function looks correct for me.  And in both 
cases the result should be the same.  Am I missing something?

-- 
regards,
Vladislav Odintsov

_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to