Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 11 Jun 2018 11:18:38 +0000
From:      bugzilla-noreply@freebsd.org
To:        net@FreeBSD.org
Subject:   [Bug 228802] [aarch64] network throughput regression in 12.0-CURRENT
Message-ID:  <bug-228802-7501-AJxdWDGBEC@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-228802-7501@https.bugs.freebsd.org/bugzilla/>
References:  <bug-228802-7501@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=3D228802

--- Comment #19 from hlh@restart.be ---
(In reply to Eugene Grosbein from comment #16)

Test are run on the same piece of hardware: boot rpool/ROOT/current (r33443=
6)
and rpool/ROOT/default (r320599).

There is a time management problem with r320599 _and_ r334436:

egrep  'spike_detect|clock_step' /var/log/ntp.log

15 Apr 09:47:50 ntpd[34364]: 0.0.0.0 0613 03 spike_detect -178.950279 s
15 Apr 09:47:51 ntpd[34364]: 0.0.0.0 061c 0c clock_step -178.950180 s
28 Apr 07:42:57 ntpd[39486]: 0.0.0.0 0613 03 spike_detect -178.959592 s
28 Apr 08:17:50 ntpd[39486]: 0.0.0.0 061c 0c clock_step -178.960942 s
29 Apr 07:50:51 ntpd[1811]: 0.0.0.0 0613 03 spike_detect -178.956145 s
29 Apr 08:01:38 ntpd[1811]: 0.0.0.0 061c 0c clock_step -178.958064 s
 2 May 13:40:24 ntpd[31174]: 0.0.0.0 0613 03 spike_detect -178.954557 s
 2 May 13:40:49 ntpd[31174]: 0.0.0.0 061c 0c clock_step -178.954612 s
 2 May 16:56:36 ntpd[31174]: 0.0.0.0 0613 03 spike_detect -357.918704 s
 2 May 17:09:24 ntpd[31174]: 0.0.0.0 061c 0c clock_step -357.918942 s
 2 May 20:34:21 ntpd[4932]: 0.0.0.0 c61c 0c clock_step -178.957489 s
 6 May 23:02:35 ntpd[20790]: 0.0.0.0 0613 03 spike_detect -178.955371 s
 6 May 23:36:10 ntpd[20790]: 0.0.0.0 061c 0c clock_step -178.955510 s
 7 May 04:50:38 ntpd[20790]: 0.0.0.0 0613 03 spike_detect -178.959438 s
 7 May 05:25:50 ntpd[20790]: 0.0.0.0 061c 0c clock_step -178.958614 s
 9 May 10:19:27 ntpd[52422]: 0.0.0.0 c61c 0c clock_step -178.957618 s
 9 May 16:48:19 ntpd[1046]: 0.0.0.0 c61c 0c clock_step -1.479122 s
13 May 04:47:30 ntpd[26110]: 0.0.0.0 0613 03 spike_detect -178.954317 s
13 May 05:07:12 ntpd[26110]: 0.0.0.0 061c 0c clock_step -178.952353 s
16 May 03:43:56 ntpd[51054]: 0.0.0.0 0613 03 spike_detect -178.958443 s
16 May 03:56:20 ntpd[51054]: 0.0.0.0 061c 0c clock_step -178.958603 s
20 May 11:08:27 ntpd[91061]: 0.0.0.0 0613 03 spike_detect -178.956436 s
20 May 11:19:35 ntpd[91061]: 0.0.0.0 061c 0c clock_step -178.957825 s
20 May 23:01:53 ntpd[88991]: 0.0.0.0 0613 03 spike_detect -178.953364 s
20 May 23:04:28 ntpd[88991]: 0.0.0.0 061c 0c clock_step -178.953372 s
22 May 09:13:28 ntpd[10332]: 0.0.0.0 0613 03 spike_detect -178.958062 s
22 May 09:19:54 ntpd[10332]: 0.0.0.0 061c 0c clock_step -178.957215 s
 2 Jun 10:35:09 ntpd[25713]: 0.0.0.0 0613 03 spike_detect -178.955744 s
 2 Jun 10:44:25 ntpd[25713]: 0.0.0.0 061c 0c clock_step -178.956775 s
 3 Jun 03:02:01 ntpd[25713]: 0.0.0.0 0613 03 spike_detect -178.953073 s
 3 Jun 03:12:24 ntpd[25713]: 0.0.0.0 061c 0c clock_step -178.953418 s
 7 Jun 11:32:29 ntpd[1842]: 0.0.0.0 c61c 0c clock_step +73.709562 s
 7 Jun 18:04:57 ntpd[1031]: 0.0.0.0 c61c 0c clock_step +63.829980 s
 8 Jun 11:13:28 ntpd[1031]: 0.0.0.0 0613 03 spike_detect -178.959113 s
 8 Jun 11:40:17 ntpd[1031]: 0.0.0.0 061c 0c clock_step -178.958836 s
 8 Jun 16:36:02 ntpd[1042]: 0.0.0.0 c61c 0c clock_step +11.228134 s
 9 Jun 11:08:50 ntpd[1034]: 0.0.0.0 c61c 0c clock_step +24.488987 s
11 Jun 10:43:33 ntpd[1038]: 0.0.0.0 c61c 0c clock_step +3.852690 s

The clock_step < 30s occurs because after a power off, I must do a boot -s
and give the current time with a date command and then resume the boot in
multiuser mode.

Note the the drift is always n*178.9 seconds with n =3D 1 or 2.
This occurs if I run something like `periodic daily`.

I will do the fetch of a file of 10MB.

--=20
You are receiving this mail because:
You are on the CC list for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-228802-7501-AJxdWDGBEC>