It looks like the test ran correctly, and the failures are caused by the performance change of more than 1%. The Mellanox team would need to comment on the requirement, as each vendor establishes their expected baseline and allowed performance change.

Cheers,
Lincoln

On Thu, Jul 16, 2020 at 11:15 PM Xia, Chenbo <chenbo.xia@intel.com> wrote:

Hi,

There’s some CI errors on Mellanox-Performance:

http://mails.dpdk.org/archives/test-report/2020-July/144853.html

http://mails.dpdk.org/archives/test-report/2020-July/144880.html

I think these patches are not related to this test.

Could someone check the issue?

Best regards,

Chenbo (Troy)

Lincoln Lavoie

Senior Engineer, Broadband Technologies

21 Madbury Rd., Ste. 100, Durham, NH 03824

lylavoie@iol.unh.edu

https://www.iol.unh.edu

+1-603-674-2755 (m)