Date: Tue, 30 Jan 2018 14:40:27 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-net@FreeBSD.org Subject: [Bug 225535] Delays in TCP connection over Gigabit Ethernet connections; Regression from 6.3-RELEASE Message-ID: <bug-225535-2472-vd4IUZ6DRn@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-225535-2472@https.bugs.freebsd.org/bugzilla/> References: <bug-225535-2472@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=225535 --- Comment #13 from Aleksander Derevianko <aeder@list.ru> --- Results of tests with kern.eventtimer.periodic=0 seems very interesting: A1 computer: grep times periodic.txt | awk '{print $3 " " $4 " " $6 " " $8 " " $10;}' | sort | uniq -c 18989 send_sync 0 0 0 0 1557 send_sync 0 0 0 1 3 send_sync 0 0 0 229 352 send_sync 0 1 0 0 A2 computer root@fspa2:~/clock/new_res # grep times periodic.txt | awk '{print $3 " " $4 " " $6 " " $8 " " $10 ;}' | sort | uniq -c 11498 send_sync 0 0 0 0 9621 send_sync 0 0 0 1 3 send_sync 0 0 0 2 4 send_sync 0 0 0 230 20 send_sync 0 1 0 0 root@fspa1:~/clock/new_res # grep times periodic.txt | grep recv_data\ 229 times 2554: send_sync 0 recv_sync 0 send_data 0 recv_data 229 eval 52 sleep 19 times 7278: send_sync 0 recv_sync 0 send_data 0 recv_data 229 eval 52 sleep 19 times 10424: send_sync 0 recv_sync 0 send_data 0 recv_data 229 eval 52 sleep 19 root@fspa1:~/clock/new_res # times 18296: send_sync 0 recv_sync 0 send_data 0 recv_data 0 eval 52 sleep 248 root@fspa2:~/clock/new_res # grep times periodic.txt| grep recv_data\ 230 times 2554: send_sync 0 recv_sync 0 send_data 0 recv_data 230 eval 52 sleep 18 times 7278: send_sync 0 recv_sync 0 send_data 0 recv_data 230 eval 52 sleep 18 times 10424: send_sync 0 recv_sync 0 send_data 0 recv_data 230 eval 52 sleep 18 times 18296: send_sync 0 recv_sync 0 send_data 0 recv_data 230 eval 52 sleep 18 root@fspa2:~/clock/new_res # As you can see, in 3 cases delay was catched on both sides, and in one case - only in A2 computer. Moreover, it seems like A1 computer have 92% of all zeroes: 18989 send_sync 0 0 0 0 1557 send_sync 0 0 0 1 At the same time, A2 computer have only 54% of all zeroes: 11498 send_sync 0 0 0 0 9621 send_sync 0 0 0 1 I have found the difference in config: A1 root@fspa1:~/clock/new_res # cat /boot/loader.conf boot_verbose="YES" coretemp_load="YES" autoboot_delay="-1" loader_logo="fbsdbw" kern.ipc.semmns=350 hw.em.txd="4096" hw.em.rxd="4096" hw.em.rx_process_limit="-1" hw.em.enable_msix="0" A2 root@fspa2:~/clock/new_res # cat /boot/loader.conf console="comconsole,vidconsole" coretemp_load="YES" autoboot_delay="-1" loader_logo="fbsdbw" kern.ipc.semmns="350" hw.em.txd="4096" hw.em.rxd="4096" hw.em.rx_process_limit="-1" hw.em.enable_msix="0" Can boot_verbose="YES" or console="comconsole,vidconsole" make such difference? -- You are receiving this mail because: You are the assignee for the bug.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-225535-2472-vd4IUZ6DRn>