Re: Poor performance with stable/13 and Mellanox ConnectX-6 (mlx5)
Date: Fri, 17 Jun 2022 00:38:55 UTC
Hi, I believe you hit the nail on the head! I am now getting consistent high speeds, even higher than on Linux! Is this a problem with the scheduler? Should someone in that area of expertise be made aware of this? More importantly i guess, would this affect real world performance, these servers will be running RabbitMQ (it uses quite a bit of bandwidth) and PostgreSQL w/ replication. [root@db-01 ~]# cpuset -l 1 iperf3 -s ----------------------------------------------------------- Server listening on 5201 (test #1) ----------------------------------------------------------- [root@db-02 ~]# cpuset -l 1 iperf3 -i 1 -t 30 -c db-01 Connecting to host db-01, port 5201 [ 5] local 192.168.10.31 port 22445 connected to 192.168.10.30 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 2.17 GBytes 18.7 Gbits/sec 263 1.13 MBytes [ 5] 1.00-2.00 sec 2.11 GBytes 18.1 Gbits/sec 0 1.14 MBytes [ 5] 2.00-3.00 sec 2.46 GBytes 21.2 Gbits/sec 696 893 KBytes [ 5] 3.00-4.00 sec 2.49 GBytes 21.4 Gbits/sec 857 447 KBytes [ 5] 4.00-5.00 sec 2.36 GBytes 20.3 Gbits/sec 281 566 KBytes [ 5] 5.00-6.00 sec 2.36 GBytes 20.3 Gbits/sec 348 593 KBytes [ 5] 6.00-7.00 sec 2.28 GBytes 19.6 Gbits/sec 0 694 KBytes [ 5] 7.00-8.00 sec 2.20 GBytes 18.9 Gbits/sec 165 611 KBytes [ 5] 8.00-9.00 sec 2.26 GBytes 19.5 Gbits/sec 183 541 KBytes [ 5] 9.00-10.00 sec 2.26 GBytes 19.4 Gbits/sec 0 625 KBytes [ 5] 10.00-11.00 sec 2.27 GBytes 19.5 Gbits/sec 165 608 KBytes [ 5] 11.00-12.00 sec 2.22 GBytes 19.0 Gbits/sec 0 679 KBytes [ 5] 12.00-13.00 sec 2.22 GBytes 19.0 Gbits/sec 0 726 KBytes [ 5] 13.00-14.00 sec 2.25 GBytes 19.3 Gbits/sec 174 610 KBytes [ 5] 14.00-15.00 sec 2.19 GBytes 18.8 Gbits/sec 0 644 KBytes [ 5] 15.00-16.00 sec 2.20 GBytes 18.9 Gbits/sec 0 704 KBytes [ 5] 16.00-17.00 sec 2.18 GBytes 18.7 Gbits/sec 0 784 KBytes [ 5] 17.00-18.00 sec 2.18 GBytes 18.7 Gbits/sec 0 820 KBytes [ 5] 18.00-19.00 sec 2.19 GBytes 18.8 Gbits/sec 0 874 KBytes [ 5] 19.00-20.00 sec 2.19 GBytes 18.8 Gbits/sec 0 902 KBytes [ 5] 20.00-21.00 sec 2.23 GBytes 19.1 Gbits/sec 202 661 KBytes [ 5] 21.00-22.00 sec 2.30 GBytes 19.8 Gbits/sec 171 894 KBytes [ 5] 22.00-23.00 sec 2.33 GBytes 20.0 Gbits/sec 341 409 KBytes [ 5] 23.00-24.00 sec 2.29 GBytes 19.7 Gbits/sec 167 594 KBytes [ 5] 24.00-25.00 sec 2.28 GBytes 19.6 Gbits/sec 0 714 KBytes [ 5] 25.00-26.00 sec 2.29 GBytes 19.7 Gbits/sec 166 604 KBytes [ 5] 26.00-27.00 sec 2.26 GBytes 19.4 Gbits/sec 0 705 KBytes [ 5] 27.00-28.00 sec 1.86 GBytes 16.0 Gbits/sec 0 763 KBytes [ 5] 28.00-29.00 sec 2.25 GBytes 19.3 Gbits/sec 0 816 KBytes [ 5] 29.00-30.00 sec 2.24 GBytes 19.2 Gbits/sec 0 869 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-30.00 sec 67.4 GBytes 19.3 Gbits/sec 4179 sender [ 5] 0.00-30.00 sec 67.4 GBytes 19.3 Gbits/sec receiver iperf Done. [root@db-02 ~]# cpuset -l 1 iperf3 -i 1 -t 30 -c db-01 Connecting to host db-01, port 5201 [ 5] local 192.168.10.31 port 51998 connected to 192.168.10.30 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 1.99 GBytes 17.1 Gbits/sec 276 456 KBytes [ 5] 1.00-2.00 sec 2.00 GBytes 17.2 Gbits/sec 0 593 KBytes [ 5] 2.00-3.00 sec 2.27 GBytes 19.5 Gbits/sec 166 556 KBytes [ 5] 3.00-4.00 sec 2.31 GBytes 19.8 Gbits/sec 186 558 KBytes [ 5] 4.00-5.00 sec 2.31 GBytes 19.8 Gbits/sec 168 558 KBytes [ 5] 5.00-6.00 sec 2.29 GBytes 19.6 Gbits/sec 182 545 KBytes [ 5] 6.00-7.00 sec 2.29 GBytes 19.7 Gbits/sec 171 507 KBytes [ 5] 7.00-8.00 sec 2.29 GBytes 19.7 Gbits/sec 180 517 KBytes [ 5] 8.00-9.00 sec 2.30 GBytes 19.7 Gbits/sec 174 568 KBytes [ 5] 9.00-10.00 sec 2.28 GBytes 19.6 Gbits/sec 178 434 KBytes [ 5] 10.00-11.00 sec 2.26 GBytes 19.4 Gbits/sec 0 583 KBytes [ 5] 11.00-12.00 sec 2.31 GBytes 19.8 Gbits/sec 180 443 KBytes [ 5] 12.00-13.00 sec 2.32 GBytes 19.9 Gbits/sec 166 564 KBytes [ 5] 13.00-14.00 sec 2.35 GBytes 20.2 Gbits/sec 350 424 KBytes [ 5] 14.00-15.02 sec 1.49 GBytes 12.5 Gbits/sec 175 1.41 KBytes [ 5] 15.02-16.00 sec 1.73 GBytes 15.2 Gbits/sec 0 1.12 MBytes [ 5] 16.00-17.00 sec 2.22 GBytes 19.1 Gbits/sec 171 784 KBytes [ 5] 17.00-18.00 sec 2.34 GBytes 20.1 Gbits/sec 348 436 KBytes [ 5] 18.00-19.00 sec 2.34 GBytes 20.1 Gbits/sec 350 547 KBytes [ 5] 19.00-20.00 sec 2.31 GBytes 19.9 Gbits/sec 173 595 KBytes [ 5] 20.00-21.00 sec 2.35 GBytes 20.2 Gbits/sec 360 439 KBytes [ 5] 21.00-22.00 sec 2.41 GBytes 20.7 Gbits/sec 525 580 KBytes [ 5] 22.00-23.00 sec 2.43 GBytes 20.9 Gbits/sec 510 397 KBytes [ 5] 23.00-24.00 sec 2.38 GBytes 20.4 Gbits/sec 532 533 KBytes [ 5] 24.00-25.00 sec 2.37 GBytes 20.4 Gbits/sec 344 547 KBytes [ 5] 25.00-26.00 sec 2.36 GBytes 20.2 Gbits/sec 354 389 KBytes [ 5] 26.00-27.00 sec 2.30 GBytes 19.8 Gbits/sec 165 592 KBytes [ 5] 27.00-28.00 sec 2.30 GBytes 19.8 Gbits/sec 173 584 KBytes [ 5] 28.00-29.00 sec 2.27 GBytes 19.5 Gbits/sec 0 701 KBytes [ 5] 29.00-30.00 sec 2.29 GBytes 19.7 Gbits/sec 0 790 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-30.00 sec 67.4 GBytes 19.3 Gbits/sec 6557 sender [ 5] 0.00-30.00 sec 67.4 GBytes 19.3 Gbits/sec receiver iperf Done. Thank You! ---- On Thu, 16 Jun 2022 17:00:25 -0400 Alexander V. Chernikov <melifaro@ipfw.ru> wrote ---- > On 16 Jun 2022, at 21:48, Mike Jakubik <mailto:mike.jakubik@swiftsmsgateway.com> wrote: > > After multiple tests and tweaks i believe the issue is not with the HW or Numa related (Infinity fabric should do around 32GB) but rather with FreeBSD TCP/IP stack. It's like it cant figure itself out properly for the speed that the HW can do, i keep getting widely varying results when testing. Below is an example of two tests, with about a 15 second break in between the two. Does pinning iperf to a specific CPU core (or range) address the variety part? e.g. cpuset -l 1 perf … The output you shared above shows CPU#83 as the core iperf is running on. Just wondering whether the scheduler migrates iperf too often, thrashing the caches.