mail ignored
2010-Jan-13 02:14 UTC
[Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On a selection of boxes, ntpd running in Xen Dom0 reproducibly exhibits extermely high noise/jitter. Switching back to -default, non-xen kernel ntpd runs with very low jitter/noise. Question -- how can I ''tame'' ntpd noise & jitter when running in Dom0? Is the problem a config issue, or a bug? Already reported this downstream; everybody''s "stumped". Running uname -a Linux test 2.6.31.8-0.1 #1 SMP 2009-12-15 23:55:40 +0100 x86_64 x86_64 x86_64 GNU/Linux with @ grub, kernel /vmlinuz root=/dev/V0/ROOT rootfstype=ext4 rootflags=journal_checksum resume=/dev/V0/SWAP showopts splash=silent selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 console=ttyS0,57600n8 cat /sys/devices/system/clocksource/clocksource0/current_clocksource tsc on 5 different mobos, all with AMD Phenom II X4 Deneb Cores (http://en.wikipedia.org/wiki/List_of_AMD_Phenom_microprocessors#.22Deneb.22_.28C2.2C_45_nm.2C_quad-core.29) e.g., grep "model name" /proc/cpuinfo model name : AMD Phenom(tm) II X4 920 Processor grep flags /proc/cpuinfo flags : fpu de tsc msr pae mce cx8 apic mtrr mca cmov pat clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt lm 3dnowext 3dnow constant_tsc rep_good nonstop_tsc pni cx16 popcnt hypervisor lahf_lm cmp_legacy extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch running ntpd, within ~ 5 minutes of launch, ntpq -p -c rv returns expected low-jitter, low-offset, full-reach results, remote refid st t when poll reach delay offset jitter ============================================================================= *clock.fmt.he.ne .PPS. 1 u 2 128 377 14.472 6.176 21.235 +otc2.psu.edu 128.4.1.1 2 u 68 128 377 98.808 17.478 33.898 +clock.isc.org 216.200.141.8 2 u 67 128 377 15.491 -4.421 22.236 +clock.sjc.he.ne .CDMA. 1 u 66 128 377 14.973 14.981 23.539 +zorro.sf-bay.or 216.218.254.202 2 u 77 128 377 14.861 -5.981 14.786 +rrcs-64-183-56- .GPS. 1 u 124 128 377 34.705 11.295 36.253 +nist1.aol-ca.tr .ACTS. 1 u 62 128 377 16.302 8.190 24.812 rebooting to, uname -a Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 x86_64 x86_64 GNU/Linux with @ grub, kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 cpufreq=xen:performance cpuidle module /vmlinuz-xen root=/dev/V0/ROOT rootfstype=ext4 rootflags=journal_checksum resume=/dev/V0/SWAP showopts splash=silent selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 console=xvc0,57600 iommu=off noirqdebug clocksource=xen and, cat /sys/devices/system/clocksource/clocksource0/current_clocksource xen cat /etc/init.d/boot.local xenpm set-scaling-governor ondemand xenpm set-scaling-governor performance after boot, cat /proc/cpuinfo | grep MHz cpu MHz : 2800.000 cpu MHz : 2800.000 cpu MHz : 2800.000 cpu MHz : 2800.000 xenpm get-cpufreq-para | grep governor current_governor : performance current_governor : performance current_governor : performance current_governor : performance ntpd @ Dom0 return extremely high offset & jitter, never dropping to ''normal'', non-xen levels; e.g., even after 2-3 days of running, remote refid st t when poll reach delay offset jitter ============================================================================= +clock.fmt.he.ne .PPS. 1 u 60 64 377 21.185 -1690.8 658.866 +otc2.psu.edu 128.4.1.1 2 u 33 64 377 109.992 -1754.1 659.825 +clock.isc.org 216.200.141.8 2 u 2 64 377 13.304 -1827.8 676.431 *clock.sjc.he.ne .CDMA. 1 u 46 64 377 21.193 -1724.5 655.151 zorro.sf-bay.or 216.218.254.202 2 u 53 64 377 15.396 -1711.7 653.852 +rrcs-64-183-56- .GPS. 1 u 28 64 377 53.359 -1766.9 657.547 +nist1.aol-ca.tr .ACTS. 1 u 37 64 377 22.613 -1742.4 665.713 Also, with ntpd *off*, @ boot to kernel-default sntp clock.isc.org && date && hwclock --show 2009 Dec 28 10:38:49.000867 + 0.250226 +/- 0.023099 secs Mon Dec 28 10:38:50 PST 2009 Mon 28 Dec 2009 10:38:51 AM PST -0.998094 seconds @ boot to kernel-xen sntp clock.isc.org && date && hwclock --show ! 2009 Dec 28 10:31:32.000206 + 28800.725203 +/- 0.013941 secs ! Mon Dec 28 18:31:32 PST 2009 Mon 28 Dec 2009 10:31:33 AM PST -0.638684 seconds note the dramatic differences in ''sntp'' & ''date'' results. These results can be reproduced swithicn back and forth between -xen & -default kernel boots. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 07:13 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Tue, Jan 12, 2010 at 06:14:57PM -0800, mail ignored wrote:> On a selection of boxes, ntpd running in Xen Dom0 reproducibly > exhibits extermely high noise/jitter. > Switching back to -default, non-xen kernel ntpd runs with very low jitter/noise. > > Question -- how can I ''tame'' ntpd noise & jitter when running in Dom0? > Is the problem a config issue, or a bug? > > Already reported this downstream; everybody''s "stumped". >How many vcpus does your dom0 have? Have you configured domain weights (ie. guaranteed cpu time for dom0?) Have you tried dedicating a single cpu core only for dom0? -- Pasi> Running > > uname -a > Linux test 2.6.31.8-0.1 #1 SMP 2009-12-15 23:55:40 +0100 x86_64 > x86_64 x86_64 GNU/Linux > > with @ grub, > > kernel /vmlinuz root=/dev/V0/ROOT rootfstype=ext4 > rootflags=journal_checksum resume=/dev/V0/SWAP showopts splash=silent > selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 > console=ttyS0,57600n8 > > cat /sys/devices/system/clocksource/clocksource0/current_clocksource > tsc > > on 5 different mobos, all with AMD Phenom II X4 Deneb Cores > (http://en.wikipedia.org/wiki/List_of_AMD_Phenom_microprocessors#.22Deneb.22_.28C2.2C_45_nm.2C_quad-core.29) > > e.g., > > grep "model name" /proc/cpuinfo > model name : AMD Phenom(tm) II X4 920 Processor > grep flags /proc/cpuinfo > flags : fpu de tsc msr pae mce cx8 apic mtrr mca cmov pat > clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt lm 3dnowext > 3dnow constant_tsc rep_good nonstop_tsc pni cx16 popcnt hypervisor > lahf_lm cmp_legacy extapic cr8_legacy abm sse4a misalignsse > 3dnowprefetch > > running ntpd, within ~ 5 minutes of launch, > > ntpq -p -c rv > > returns expected low-jitter, low-offset, full-reach results, > > remote refid st t when poll reach delay offset > jitter > =============================================================================> *clock.fmt.he.ne .PPS. 1 u 2 128 377 14.472 6.176 > 21.235 > +otc2.psu.edu 128.4.1.1 2 u 68 128 377 98.808 17.478 > 33.898 > +clock.isc.org 216.200.141.8 2 u 67 128 377 15.491 -4.421 > 22.236 > +clock.sjc.he.ne .CDMA. 1 u 66 128 377 14.973 14.981 > 23.539 > +zorro.sf-bay.or 216.218.254.202 2 u 77 128 377 14.861 -5.981 > 14.786 > +rrcs-64-183-56- .GPS. 1 u 124 128 377 34.705 11.295 > 36.253 > +nist1.aol-ca.tr .ACTS. 1 u 62 128 377 16.302 8.190 > 24.812 > > rebooting to, > > uname -a > Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 > x86_64 x86_64 GNU/Linux > > with @ grub, > > kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M > vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 > cpufreq=xen:performance cpuidle > module /vmlinuz-xen root=/dev/V0/ROOT rootfstype=ext4 > rootflags=journal_checksum resume=/dev/V0/SWAP showopts splash=silent > selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 > console=xvc0,57600 iommu=off noirqdebug clocksource=xen > > and, > > cat /sys/devices/system/clocksource/clocksource0/current_clocksource > xen > > cat /etc/init.d/boot.local > xenpm set-scaling-governor ondemand > xenpm set-scaling-governor performance > > after boot, > > cat /proc/cpuinfo | grep MHz > cpu MHz : 2800.000 > cpu MHz : 2800.000 > cpu MHz : 2800.000 > cpu MHz : 2800.000 > > xenpm get-cpufreq-para | grep governor > current_governor : performance > current_governor : performance > current_governor : performance > current_governor : performance > > ntpd @ Dom0 return extremely high offset & jitter, never dropping to > ''normal'', non-xen levels; e.g., even after 2-3 days of running, > > remote refid st t when poll reach delay offset jitter > =============================================================================> +clock.fmt.he.ne .PPS. 1 u 60 64 377 21.185 -1690.8 > 658.866 > +otc2.psu.edu 128.4.1.1 2 u 33 64 377 109.992 -1754.1 > 659.825 > +clock.isc.org 216.200.141.8 2 u 2 64 377 13.304 -1827.8 > 676.431 > *clock.sjc.he.ne .CDMA. 1 u 46 64 377 21.193 -1724.5 > 655.151 > zorro.sf-bay.or 216.218.254.202 2 u 53 64 377 15.396 -1711.7 > 653.852 > +rrcs-64-183-56- .GPS. 1 u 28 64 377 53.359 -1766.9 > 657.547 > +nist1.aol-ca.tr .ACTS. 1 u 37 64 377 22.613 -1742.4 > 665.713 > > Also, with ntpd *off*, > > @ boot to kernel-default > sntp clock.isc.org && date && hwclock --show > 2009 Dec 28 10:38:49.000867 + 0.250226 +/- 0.023099 secs > Mon Dec 28 10:38:50 PST 2009 > Mon 28 Dec 2009 10:38:51 AM PST -0.998094 seconds > > @ boot to kernel-xen > sntp clock.isc.org && date && hwclock --show > ! 2009 Dec 28 10:31:32.000206 + 28800.725203 +/- 0.013941 secs > ! Mon Dec 28 18:31:32 PST 2009 > Mon 28 Dec 2009 10:31:33 AM PST -0.638684 seconds > > note the dramatic differences in ''sntp'' & ''date'' results. These > results can be reproduced swithicn back and forth between -xen & > -default kernel boots. > > _______________________________________________ > Xen-users mailing list > Xen-users@lists.xensource.com > http://lists.xensource.com/xen-users_______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-13 07:37 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 9:14 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> On a selection of boxes, ntpd running in Xen Dom0 reproducibly > exhibits extermely high noise/jitter. > Switching back to -default, non-xen kernel ntpd runs with very low jitter/noise.> uname -a > Linux test 2.6.31.8-0.1 #1 SMP 2009-12-15 23:55:40 +0100 x86_64 > x86_64 x86_64 GNU/Linuxwhich kernel is it? vanilla? distro-provided? gentoo-xen kernel? FWIW, a quick workaround is to use kernel 2.6.18. I''ve been using RHEL''s kernel-xen (2.6.18-164.2.1.el5xen) and it works great WRT time/ntp. -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 16:44 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi On Tue, Jan 12, 2010 at 11:13 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> How many vcpus does your dom0 have?4, atm, xm vcpu-list Domain-0 Name ID VCPU CPU State Time(s) CPU Affinity Domain-0 0 0 0 r-- 273.0 any cpu Domain-0 0 1 1 -b- 202.4 any cpu Domain-0 0 2 3 -b- 179.1 any cpu Domain-0 0 3 2 -b- 199.9 any cpu> Have you configured domain weights (ie. guaranteed cpu time for dom0?)no, i didn''t realize that was available. how is that explicitly done? is that done using the CreditScheduler?> Have you tried dedicating a single cpu core only for dom0?no, not yet. clearly, i could set maxcpus=1 @ grub. would i also need to pin the Dom0 cpu to a specific cpu, shomehow? On Tue, Jan 12, 2010 at 11:37 PM, Fajar A. Nugraha <fajar@fajar.net> wrote:> which kernel is it? vanilla? distro-provided? gentoo-xen kernel?it''s distro-provided -- opensuse 11.2''s, uname -a Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 x86_64 x86_64 GNU/Linux lsb_release -a LSB Version: core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:graphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:graphics-4.0-noarch Distributor ID: SUSE LINUX Description: openSUSE 11.2 (x86_64) Release: 11.2 Codename: n/a rpm -qa | grep -i ^xen-3 xen-3.4.1_19718_04-28.1.x86_64> FWIW, a quick workaround is to use kernel 2.6.18. I''ve been using > RHEL''s kernel-xen (2.6.18-164.2.1.el5xen) and it works great WRT > time/ntp.good to know. although @downstream, they''re unable to reproduce this problem, i''m able to on, as above, 5 different mobos from 3 different vendors. commonalities are (1) opensuse 11.2, (2) similar install profiles, (3) AMD Deneb cores (Phenom II X4s) whether this is an issue due to my local environmentS, opensuse or xen, i''d like to find/fix it. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 17:13 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 08:44:08AM -0800, mail ignored wrote:> hi > > On Tue, Jan 12, 2010 at 11:13 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > > How many vcpus does your dom0 have? > > 4, atm, > > xm vcpu-list Domain-0 > Name ID VCPU CPU State Time(s) CPU Affinity > Domain-0 0 0 0 r-- 273.0 any cpu > Domain-0 0 1 1 -b- 202.4 any cpu > Domain-0 0 2 3 -b- 179.1 any cpu > Domain-0 0 3 2 -b- 199.9 any cpu > > > Have you configured domain weights (ie. guaranteed cpu time for dom0?) > > no, i didn''t realize that was available. how is that explicitly done? > is that done using the CreditScheduler? >Yes, it''s done using Xen Credit Scheduler: http://wiki.xensource.com/xenwiki/CreditScheduler Use this to check the current settings: xm sched-credit -d Domain-0 And this to configure the weight: xm sched-credit -d Domain-0 -w <weight> You might want to give dom0 for example weight of 384 or 512, so it''ll have more weight than the guests (the default weight is 256).> > Have you tried dedicating a single cpu core only for dom0? > > no, not yet. clearly, i could set maxcpus=1 @ grub. would i also > need to pin the Dom0 cpu to a specific cpu, shomehow? >You can use this for xen.gz in grub.conf: dom0_vcpus=1 dom0_vcpus_pin And then make sure the other domains (guests) don''t use pcpu 0, using cpus=1-3 parameter in /etc/xen/<guest> cfgfiles. Also you might want to dedicate dom0 fixed amount of memory, for example 512MB using dom0_mem=512M options for xen.gz, and then disable dom0 ballooning totally by specifying (dom0-min-mem 512) in /etc/xen/xend-config.sxp -- Pasi> On Tue, Jan 12, 2010 at 11:37 PM, Fajar A. Nugraha <fajar@fajar.net> wrote: > > which kernel is it? vanilla? distro-provided? gentoo-xen kernel? > > it''s distro-provided -- opensuse 11.2''s, > > uname -a > Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 > x86_64 x86_64 GNU/Linux > > lsb_release -a > LSB Version: > core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:graphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:graphics-4.0-noarch > Distributor ID: SUSE LINUX > Description: openSUSE 11.2 (x86_64) > Release: 11.2 > Codename: n/a > > rpm -qa | grep -i ^xen-3 > xen-3.4.1_19718_04-28.1.x86_64 > > > FWIW, a quick workaround is to use kernel 2.6.18. I''ve been using > > RHEL''s kernel-xen (2.6.18-164.2.1.el5xen) and it works great WRT > > time/ntp. > > good to know. although @downstream, they''re unable to reproduce this > problem, i''m able to on, as above, 5 different mobos from 3 different > vendors. > > commonalities are (1) opensuse 11.2, (2) similar install profiles, (3) > AMD Deneb cores (Phenom II X4s) > > whether this is an issue due to my local environmentS, opensuse or > xen, i''d like to find/fix it._______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 17:23 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi On Wed, Jan 13, 2010 at 9:13 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> Yes, it''s done using Xen Credit Scheduler: > http://wiki.xensource.com/xenwiki/CreditScheduleralready reading up!> You might want to give dom0 for example weight of 384 or 512, > so it''ll have more weight than the guests (the default weight is 256).missed the default value. that explains that. thanks.> dom0_vcpus=1 dom0_vcpus_pindoes that *presume* pin to cpu=0? iiuc, i *should* be also able to dom0_vcpus=2 dom0_vcpus_pin and specify that it pins to cpus=0&1? or doe the pinning always use the lowest CPU #s?> And then make sure the other domains (guests) don''t use pcpu 0, using > cpus=1-3 parameter in /etc/xen/<guest> cfgfiles.got it.> Also you might want to dedicate dom0 fixed amount of memory, for example > 512MB using dom0_mem=512M options for xen.gz, and then disable dom0 > ballooning totally by specifying (dom0-min-mem 512) in /etc/xen/xend-config.sxpthat''s already done in my case, @ 1024. thanks! _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 17:27 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
forgot to ask ...> And this to configure the weight: > xm sched-credit -d Domain-0 -w <weight>where/what is the "right" place/method to set those for survive across boot? grub? in /etc/init.d/(late boot script)?> You might want to give dom0 for example weight of 384 or 512, > so it''ll have more weight than the guests (the default weight is 256).> You can use this for xen.gz in grub.conf: > dom0_vcpus=1 dom0_vcpus_pinhow is maxcpus=N different? is that a ''normal'' kernel construct? do i still need it, if using dom0_vcpus? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 17:37 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 09:27:22AM -0800, mail ignored wrote:> forgot to ask ... > > > And this to configure the weight: > > xm sched-credit -d Domain-0 -w <weight> > > where/what is the "right" place/method to set those for survive across boot? > > grub? in /etc/init.d/(late boot script)? >/etc/rc.local maybe.. you need to do it after xend has started.> > You might want to give dom0 for example weight of 384 or 512, > > so it''ll have more weight than the guests (the default weight is 256). > > > You can use this for xen.gz in grub.conf: > > dom0_vcpus=1 dom0_vcpus_pin > > how is maxcpus=N different? is that a ''normal'' kernel construct? do > i still need it, if using dom0_vcpus? >you don''t need Linux kernel maxcpus= option when you specify Xen hypervisor dom0_vcpus=X. Dom0 Linux kernel won''t see more than X CPUs then. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 17:37 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 09:23:46AM -0800, mail ignored wrote:> > > You might want to give dom0 for example weight of 384 or 512, > > so it''ll have more weight than the guests (the default weight is 256). > > missed the default value. that explains that. thanks. > > > dom0_vcpus=1 dom0_vcpus_pin > > does that *presume* pin to cpu=0? > > iiuc, i *should* be also able to > > dom0_vcpus=2 dom0_vcpus_pin > > and specify that it pins to cpus=0&1? or doe the pinning always use > the lowest CPU #s? >Yeah, "dom0_vcpus=2 dom0_vcpus_pin" should work. You can verify it by running "xm vcpu-list" after reboot. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 17:42 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
ok, almost there.> On Wed, Jan 13, 2010 at 9:13 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: >> And then make sure the other domains (guests) don''t use pcpu 0, using >> cpus=1-3 parameter in /etc/xen/<guest> cfgfiles.that goes on the extra=''... cpus=1-3 ...'' line, yes? as a standalone entry in the cfg, it''s certainly not happy ... _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 17:45 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 09:42:53AM -0800, mail ignored wrote:> ok, almost there. > > > On Wed, Jan 13, 2010 at 9:13 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > >> And then make sure the other domains (guests) don''t use pcpu 0, using > >> cpus=1-3 parameter in /etc/xen/<guest> cfgfiles. > > that goes on the extra=''... cpus=1-3 ...'' line, yes? > > as a standalone entry in the cfg, it''s certainly not happy ... >No, it doesn''t go to extra. cpus="1-3" vcpus=1 That should do it. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 17:51 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 9:45 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> cpus="1-3"agh! quotes ... now to set it all up, and watch ntpq for a bit ... thanks. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 18:06 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 9:13 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> You can use this for xen.gz in grub.conf: > dom0_vcpus=1 dom0_vcpus_pinbooting to, kernel /xen.gz ... dom0_vcpus=1 dom0_vcpus_pin sched=credit ... i check, xm list Name ID Mem VCPUs State Time(s) Domain-0 0 1022 4 r----- 79.4 xm vcpu-list Domain-0 Name ID VCPU CPU State Time(s) CPU Affinity Domain-0 0 0 0 -b- 37.4 0 Domain-0 0 1 1 -b- 13.0 1 Domain-0 0 2 2 r-- 21.6 2 Domain-0 0 3 3 -b- 15.1 3 but, xm vcpu-set --help Usage: xm vcpu-set <Domain> <vCPUs> Set the number of active VCPUs for allowed for the domain. then, xm vcpu-set Domain-0 1 xm vcpu-list Domain-0 (locks up the ssh session ... no response) ssh in another session, try again, xm vcpu-list Domain-0 (locks up the ssh session ... no response) so, not sticking across boot ... and this odd lockup. thoughts? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 18:08 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:06:26AM -0800, mail ignored wrote:> On Wed, Jan 13, 2010 at 9:13 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > > You can use this for xen.gz in grub.conf: > > dom0_vcpus=1 dom0_vcpus_pin > > booting to, > > kernel /xen.gz ... dom0_vcpus=1 dom0_vcpus_pin sched=credit ... > > i check, > > xm list > Name ID Mem VCPUs > State Time(s) > Domain-0 0 1022 4 r----- 79.4 > > xm vcpu-list Domain-0 > Name ID VCPU CPU State Time(s) > CPU Affinity > Domain-0 0 0 0 -b- 37.4 0 > Domain-0 0 1 1 -b- 13.0 1 > Domain-0 0 2 2 r-- 21.6 2 > Domain-0 0 3 3 -b- 15.1 3 >.. so something went wrong. You should only see 1 vcpus for dom0 now.> but, > > xm vcpu-set --help > Usage: xm vcpu-set <Domain> <vCPUs> > > Set the number of active VCPUs for allowed for the domain. > > then, > > xm vcpu-set Domain-0 1 > xm vcpu-list Domain-0 > > (locks up the ssh session ... no response) > > ssh in another session, try again, > > xm vcpu-list Domain-0 > > (locks up the ssh session ... no response) > > > so, not sticking across boot ... and this odd lockup. thoughts? >What dom0 kernel you were running again? I can''t remember anymore. Looks like it didn''t like when cpus were removed on-the-fly. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 18:10 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:08 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> What dom0 kernel you were running again? I can''t remember anymore.np. it''s distro-provided -- opensuse 11.2''s, uname -a Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 x86_64 x86_64 GNU/Linux lsb_release -a LSB Version: core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:graphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:graphics-4.0-noarch Distributor ID: SUSE LINUX Description: openSUSE 11.2 (x86_64) Release: 11.2 Codename: n/a rpm -qa | grep -i ^xen-3 xen-3.4.1_19718_04-28.1.x86_64> Looks like it didn''t like when cpus were removed on-the-fly.which is fine, i suppose, as long as we can get it to stick at boot. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 18:12 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:10:32AM -0800, mail ignored wrote:> On Wed, Jan 13, 2010 at 10:08 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > > What dom0 kernel you were running again? I can''t remember anymore. > > np. > > it''s distro-provided -- opensuse 11.2''s, >OK.. so you might want to report a bug to the opensuse bugzilla.> uname -a > Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 > x86_64 x86_64 GNU/Linux > > lsb_release -a > LSB Version: > core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:graphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:graphics-4.0-noarch > Distributor ID: SUSE LINUX > Description: openSUSE 11.2 (x86_64) > Release: 11.2 > Codename: n/a > > rpm -qa | grep -i ^xen-3 > xen-3.4.1_19718_04-28.1.x86_64 > > > Looks like it didn''t like when cpus were removed on-the-fly. > > which is fine, i suppose, as long as we can get it to stick at boot. >"dom0_vcpus=1 dom0_vcpus=pin" options should do it. Verify with "xm dmesg" that there parameters are really there. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 18:25 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:12 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> OK.. so you might want to report a bug to the opensuse bugzilla.so it *is* a bug. ok, i''ll file @ opensuse. thanks.> "dom0_vcpus=1 dom0_vcpus=pin" options should do it. > Verify with "xm dmesg" that there parameters are really there.seem to be, xm dmesg | egrep -i "dom0|vcp" (XEN) Command line: loglvl=all loglvl_guest=all dom0_mem=1024M vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 cpufreq=xen:performance cpuidle dom0_vcpus=1 dom0_vcpus_pin sched=credit (XEN) Dom0 kernel: 64-bit, lsb, paddr 0x2000 -> 0x8e1000 (XEN) Dom0 alloc.: 0000000216000000->0000000218000000 (253952 pages to be allocated) (XEN) Dom0 has maximum 4 VCPUs (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch input to Xen) but it is, then. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 18:39 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:25:45AM -0800, mail ignored wrote:> On Wed, Jan 13, 2010 at 10:12 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > > OK.. so you might want to report a bug to the opensuse bugzilla. > > so it *is* a bug. ok, i''ll file @ opensuse. thanks. >I think it''s actually two different bugs :) 1) You should be able to limit the number of dom0 vcpus by dom0_vcpus option 2) It shouldn''t crash when you remove vcpus from dom0 using vcpu-set.> > "dom0_vcpus=1 dom0_vcpus=pin" options should do it. > > Verify with "xm dmesg" that there parameters are really there. > > seem to be, > > xm dmesg | egrep -i "dom0|vcp" > (XEN) Command line: loglvl=all loglvl_guest=all dom0_mem=1024M > vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 > cpufreq=xen:performance cpuidle dom0_vcpus=1 dom0_vcpus_pin > sched=credit > (XEN) Dom0 kernel: 64-bit, lsb, paddr 0x2000 -> 0x8e1000 > (XEN) Dom0 alloc.: 0000000216000000->0000000218000000 (253952 > pages to be allocated) > (XEN) Dom0 has maximum 4 VCPUs > (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch > input to Xen) > > but it is, then. >I''ll test dom0_vcpus option on my testbox at some point, with Xen 3.4.x. I''ve been using it successfully with Xen 3.1.x, on RHEL5/CentOS5. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 18:58 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:39 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> I think it''s actually two different bugs :) > > 1) You should be able to limit the number of dom0 vcpus by dom0_vcpus option > 2) It shouldn''t crash when you remove vcpus from dom0 using vcpu-set.thanks. two bugs filed. fyi, "Bug 570432 - xen ignores cmd line params to limit/pin Dom0 cpu utilization" https://bugzilla.novell.com/show_bug.cgi?id=570432 "Bug 570443 - CRASH @ attempt to manually remove vcpus from dom0 using vcpu-set" https://bugzilla.novell.com/show_bug.cgi?id=570443> I''ll test dom0_vcpus option on my testbox at some point, with Xen 3.4.x. > I''ve been using it successfully with Xen 3.1.x, on RHEL5/CentOS5.would be helpful to know. thanks. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 20:38 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 08:39:40PM +0200, Pasi Kärkkäinen wrote:> On Wed, Jan 13, 2010 at 10:25:45AM -0800, mail ignored wrote: > > On Wed, Jan 13, 2010 at 10:12 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > > > OK.. so you might want to report a bug to the opensuse bugzilla. > > > > so it *is* a bug. ok, i''ll file @ opensuse. thanks. > > > > I think it''s actually two different bugs :) > > 1) You should be able to limit the number of dom0 vcpus by dom0_vcpus option > 2) It shouldn''t crash when you remove vcpus from dom0 using vcpu-set. > > > > "dom0_vcpus=1 dom0_vcpus=pin" options should do it. > > > Verify with "xm dmesg" that there parameters are really there. > > > > seem to be, > > > > xm dmesg | egrep -i "dom0|vcp" > > (XEN) Command line: loglvl=all loglvl_guest=all dom0_mem=1024M > > vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 > > cpufreq=xen:performance cpuidle dom0_vcpus=1 dom0_vcpus_pin > > sched=credit > > (XEN) Dom0 kernel: 64-bit, lsb, paddr 0x2000 -> 0x8e1000 > > (XEN) Dom0 alloc.: 0000000216000000->0000000218000000 (253952 > > pages to be allocated) > > (XEN) Dom0 has maximum 4 VCPUs > > (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch > > input to Xen) > > > > but it is, then. > > > > I''ll test dom0_vcpus option on my testbox at some point, with Xen 3.4.x. > I''ve been using it successfully with Xen 3.1.x, on RHEL5/CentOS5. >I just tried it on my Fedora 12 testbox with Xen 3.4.2, and I''m seeing the same behaviour: (XEN) Xen version 3.4.2 (root@(none)) (gcc version 4.4.2 20091027 (RedHat 4.4.2-7) (GCC) ) Mon Dec 14 06:53:18 EET 2009 (XEN) Latest ChangeSet: unavailable (XEN) Command line: dom0_mem=1G loglvl=all guest_loglvl=all com1=38400,8n1 console=com1 dom0_vcpus=1 .. (XEN) Dom0 has maximum 4 VCPUs .. I just sent an email to xen-devel asking about it.. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 20:41 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 12:38 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> I just tried it on my Fedora 12 testbox with Xen 3.4.2, and I''m seeing > the same behaviour: > > (XEN) Xen version 3.4.2 (root@(none)) (gcc version 4.4.2 20091027 (RedHat 4.4.2-7) (GCC) ) Mon Dec 14 06:53:18 EET 2009 > (XEN) Latest ChangeSet: unavailable > (XEN) Command line: dom0_mem=1G loglvl=all guest_loglvl=all com1=38400,8n1 console=com1 dom0_vcpus=1 > .. > (XEN) Dom0 has maximum 4 VCPUshrm. bad to see. good to know.> I just sent an email to xen-devel asking about it..i''ll keep an eye out there, as well. thanks. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-13 21:55 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 10:38:18PM +0200, Pasi Kärkkäinen wrote:> On Wed, Jan 13, 2010 at 08:39:40PM +0200, Pasi Kärkkäinen wrote: > > On Wed, Jan 13, 2010 at 10:25:45AM -0800, mail ignored wrote: > > > On Wed, Jan 13, 2010 at 10:12 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote: > > > > OK.. so you might want to report a bug to the opensuse bugzilla. > > > > > > so it *is* a bug. ok, i''ll file @ opensuse. thanks. > > > > > > > I think it''s actually two different bugs :) > > > > 1) You should be able to limit the number of dom0 vcpus by dom0_vcpus option > > 2) It shouldn''t crash when you remove vcpus from dom0 using vcpu-set. > > > > > > "dom0_vcpus=1 dom0_vcpus=pin" options should do it. > > > > Verify with "xm dmesg" that there parameters are really there. > > > > > > seem to be, > > > > > > xm dmesg | egrep -i "dom0|vcp" > > > (XEN) Command line: loglvl=all loglvl_guest=all dom0_mem=1024M > > > vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 > > > cpufreq=xen:performance cpuidle dom0_vcpus=1 dom0_vcpus_pin > > > sched=credit > > > (XEN) Dom0 kernel: 64-bit, lsb, paddr 0x2000 -> 0x8e1000 > > > (XEN) Dom0 alloc.: 0000000216000000->0000000218000000 (253952 > > > pages to be allocated) > > > (XEN) Dom0 has maximum 4 VCPUs > > > (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch > > > input to Xen) > > > > > > but it is, then. > > > > > > > I''ll test dom0_vcpus option on my testbox at some point, with Xen 3.4.x. > > I''ve been using it successfully with Xen 3.1.x, on RHEL5/CentOS5. > > > > I just tried it on my Fedora 12 testbox with Xen 3.4.2, and I''m seeing > the same behaviour: > > (XEN) Xen version 3.4.2 (root@(none)) (gcc version 4.4.2 20091027 (RedHat 4.4.2-7) (GCC) ) Mon Dec 14 06:53:18 EET 2009 > (XEN) Latest ChangeSet: unavailable > (XEN) Command line: dom0_mem=1G loglvl=all guest_loglvl=all com1=38400,8n1 console=com1 dom0_vcpus=1 > .. > (XEN) Dom0 has maximum 4 VCPUs > .. > > I just sent an email to xen-devel asking about it.. >like Andy pointed out on xen-devel, the xen.gz option is called dom0_max_vcpus nowadays. I verified dom0_max_vcpus=1 works OK with Xen 3.4.2. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 22:10 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 1:55 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> like Andy pointed out on xen-devel, the xen.gz option is called > dom0_max_vcpus nowadays. > > I verified dom0_max_vcpus=1 works OK with Xen 3.4.2.and it does with my xen 3.4.1 on os11.2. after boot, xm dmesg | egrep -i "dom0|vcp" (XEN) Command line: loglvl=all ... dom0_max_vcpus=1 dom0_vcpus_pin sched=credit (XEN) Dom0 kernel: 64-bit, lsb, paddr 0x2000 -> 0x8e1000 (XEN) Dom0 alloc.: 0000000216000000->0000000218000000 (253952 pages to be allocated) (XEN) Dom0 has maximum 1 VCPUs (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch input to Xen) xm vcpu-list Domain-0 Name ID VCPU CPU State Time(s) CPU Affinity Domain-0 0 0 0 r-- 48.0 0 thanks for the help! now back to watching ntpq ... _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-13 23:36 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Wed, Jan 13, 2010 at 2:10 PM, mail ignored <0.bugs.only.0@gmail.com> wrote:> now back to watching ntpq ...after applying the suggestions above, we''ve still got excessive offset/jitter ... boot to, title xen root (hd0,0) kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 cpufreq=xen:performance cpuidle dom0_max_vcpus=1 dom0_vcpus_pin sched=credit module /vmlinuz-xen root=/dev/V0/ROOT rootfstype=ext4 rootflags=journal_checksum resume=/dev/V0/SWAP showopts splash=silent selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 console=xvc0,57600 iommu=off noirqdebug clocksource=xen module /initrd-xen with, cat /proc/cpuinfo | grep MHz cpu MHz : 2800.000 cpu MHz : 2800.000 cpu MHz : 2800.000 cpu MHz : 2800.000 xenpm get-cpufreq-para | grep governor current_governor : performance current_governor : performance current_governor : performance current_governor : performance cat /etc/init.d/boot.local /usr/sbin/xenpm set-scaling-governor ondemand /usr/sbin/xenpm set-scaling-governor performance /usr/sbin/xm sched-credit -d Domain-0 -w 512 for DomU == "test_guest", cat /etc/xen/vm/test_guest.cfg ... cpus = ''1-3'' vcpus = 3 localtime = 0 ... xm list Name ID Mem VCPUs State Time(s) Domain-0 0 1022 1 r----- 55.1 test_guest 1 1024 3 -b---- 7.9 xm vcpu-list Domain-0 Name ID VCPU CPU State Time(s) CPU Affinity Domain-0 0 0 0 r-- 56.6 0 xm vcpu-list test_guest Name ID VCPU CPU State Time(s) CPU Affinity test_guest 1 0 3 -b- 3.6 1-3 test_guest 1 1 2 -b- 2.2 1-3 test_guest 1 2 1 -b- 2.1 1-3 xm sched-credit -d Domain-0 Name ID Weight Cap Domain-0 0 256 0 ??? hrm. checking, cat /etc/init.d/boot.local /usr/sbin/xenpm set-scaling-governor ondemand /usr/sbin/xenpm set-scaling-governor performance /usr/sbin/xm sched-credit -d Domain-0 -w 512 xm dmesg | grep -i scheduler (XEN) Using scheduler: SMP Credit Scheduler (credit) xm sched-credit -d Domain-0 -w 512 xm sched-credit -d Domain-0 Name ID Weight Cap Domain-0 0 512 0 xm sched-credit -d test_guest Name ID Weight Cap test_guest 1 256 0 grep dom0 /etc/xen/xend-config.sxp (dom0-min-mem 0) (enable-dom0-ballooning no) (dom0-cpus 0) after ~ 1/2 hour, uptime 15:29pm up 0:34, 1 user, load average: 0.00, 0.00, 0.02 ntp sync has reached state=2 @ stratum=2, but i''m still seeing the excessive Xen offset/jitter -- only under Xen, e.g. ntpq -p -c rv assID=0 status=06d4 leap_none, sync_ntp, 13 events, event_peer/strat_chg, version="ntpd 4.2.4p8@1.1612-o Tue Dec 15 14:06:22 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.8-0.1-xen", leap=00, stratum=2, precision=-20, rootdelay=17.385, rootdispersion=125.916, peer=7611, refid=207.200.81.113, reftime=cef8d28b.247d36d5 Wed, Jan 13 2010 15:15:23.142, poll=6, clock=cef8d5e2.cd45b382 Wed, Jan 13 2010 15:29:38.801, state=2, offset=-90.707, frequency=-500.000, jitter=513.693, noise=25.184, stability=154.680, tai=0 remote refid st t when poll reach delay offset jitter ============================================================================= +clock.fmt.he.ne .PPS. 1 u 37 64 377 16.531 -1495.9 522.188 +otc2.psu.edu 129.6.15.29 2 u 24 64 377 108.729 -1514.8 516.844 +clock.isc.org .GPS. 1 u 31 64 377 20.388 -1509.1 518.560 +clock.sjc.he.ne .CDMA. 1 u 48 64 377 20.548 -1475.4 514.975 +zorro.sf-bay.or 216.218.254.202 2 u 48 64 377 19.124 -1476.2 520.580 *nist1.aol-ca.tr .ACTS. 1 u 6 64 377 26.142 -1543.1 511.533 so CPU isolation is not involved ... ? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-14 16:32 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi, On Wed, Jan 13, 2010 at 1:55 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> like Andy pointed out on xen-devel, the xen.gz option is called > dom0_max_vcpus nowadays. > > I verified dom0_max_vcpus=1 works OK with Xen 3.4.2.I''ve just been told @novell bugzilla, "Please be also advised that dom0_max_vcpus=1 is not recommended (and not supported); a minimum of 2 should be set here." Any thoughts/comments? Should it be used @ =1, or not? confusing, atm. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-15 02:19 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Thu, Jan 14, 2010 at 11:32 PM, mail ignored <0.bugs.only.0@gmail.com> wrote:> hi, > > On Wed, Jan 13, 2010 at 1:55 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote: >> like Andy pointed out on xen-devel, the xen.gz option is called >> dom0_max_vcpus nowadays. >> >> I verified dom0_max_vcpus=1 works OK with Xen 3.4.2. > > I''ve just been told @novell bugzilla, > > "Please be also advised that dom0_max_vcpus=1 is not recommended (and > not supported); a minimum of 2 should be set here." > > Any thoughts/comments? Should it be used @ =1, or not? confusing, atm.Your original problem was that ntpd experience high noise/jitter. IMHO dedicated dom0 cpu is not the fix for that, so that''s two different issues. First, about ntpd. Since you''re using Opensuse kernel, your best bet is novell guys, since users on this list does not know what cause or fix that issue. A quick workaround is to use Redhat''s kernel, which worked fine for me with both Intel and AMD cpus. But from your posts I gather you want to stick with Opensuse kernel. Second, about dom0 cpu. It''s generally best to assign at least one physical core/cpu exclusive for dom0 usage. What I did is use this on grub.conf kernel /xen.gz-3.4.1 com2=115200,8n1 console=vga,com2 dom0_mem=768M dom0_vcpus_pin and this on xend-config.sxp (dom0-cpus 1) and set domUs to use cpus other than cpu0, with a config line like "cpus=1-7" The difference between this approach and the one you''re trying to use: - I hard-limit dom0 memory. This is useful because dom0 won''t have to balloon-down memory for domU usage, plus it doesn''t have to maintain necessary memory metadata to support full physical available memory (thus more efficient memory usage) - I don''t limit dom0 cpu on xen.gz line, but I do it on xend-config.sxp. What this means is that dom0 will start initially with all available cpu, but when xend starts it changes to only 1 cpu. The part "dom_vcpus_pin" ensures that the active cpu for dom0 will be cpu0, thus you can assign other cpus to domUs. The benefit of this approach is that it allows me to later add more cpus to dom0 when necessary (for example, if I want to compile a new kernel on dom0). If I understand correctly, dom0_max_vcpus=1 means that you hardlimit dom0''s cpu to 1, thus you''ll be unable to add more cpu to dom0 without reboot. This might be the reason why novell guys recommended against it. You''ll have to ask them for the details though. Note that "unsupported" does not allways mean it won''t work :D -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 03:11 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Thu, Jan 14, 2010 at 6:19 PM, Fajar A. Nugraha <fajar@fajar.net> wrote:> Your original problem was that ntpd experience high noise/jitter. IMHO > dedicated dom0 cpu is not the fix for that, so that''s two different > issues.It certainly did not appear to make a difference.> First, about ntpd. Since you''re using Opensuse kernel, your best bet > is novell guys, since users on this list does not know what cause or > fix that issue.As per my OP, "Already reported this downstream; everybody''s "stumped". I.e., it''s already been filed at novell, and discussed ad nauseum. No solution/progress there, simply statements that they can''t reproduce the problem; of course, that''s with not identical environments. I *am* able to reproduce this issue on 5 different boxes, so I''m hard pressed to believe that its any one box that''s at fault., There have been suggestions to come ''here'', though. And, there are @novell types on these lists -- if more prominently on xen-devel> Second, about dom0 cpu. It''s generally best to assign at least one > physical core/cpu exclusive for dom0 usage. What I did is use this on > grub.conf > > kernel /xen.gz-3.4.1 com2=115200,8n1 console=vga,com2 dom0_mem=768M > dom0_vcpus_pin...> The difference between this approach and the one you''re trying to use: > - I hard-limit dom0 memory.you''ll note from above that i already *have*limited dom0 memory, in grub, kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 iommu=1 cpufreq=xen:performance cpuidle dom0_max_vcpus=2 dom0_vcpus_pin sched=credit have already disabled ballooning in xend-config.sxp, (enable-dom0-ballooning no) (dom0-cpus 0), and, have already isolated guests to the ''other'' cpus, e.g., in test_guest.cfg ... cpus = ''2-3'' vcpus = 2 ... with this config I verify 2 cpus available to Dom0 cat /proc/cpuinfo | grep processor processor : 0 processor : 1 with 2 vcpus using those two cpus, xm vcpu-list Domain-0 Name ID VCPU CPU State Time(s) CPU Affinity Domain-0 0 0 0 -b- 31.5 0 Domain-0 0 1 1 r-- 24.6 1 and @ the guest, the ''other'' two cpus being used, xm vcpu-list test_guest Name ID VCPU CPU State Time(s) CPU Affinity test_guest 1 0 3 -b- 5.0 2-3 test_guest 1 1 2 -b- 2.4 2-3> This might be the reason why novell guys recommended against it. > You''ll have to ask them for the details though.already have ... waiting for clarification> Note that "unsupported" does not allways mean it won''t work :Dthat''s certainly true! the point remains, however, that ntpd withOUT xen''s presence, on any/all hardware I have, runs quiet & stable. clearly, "xen" is involved. whether it''s the opensuse implementation of kernel-xen or Xen, or something inherent in Xen''s design, is unknown atm ... thanks. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-15 03:23 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 10:11 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> xm vcpu-list Domain-0 > Name ID VCPU CPU State Time(s) CPU Affinity > Domain-0 0 0 0 -b- 31.5 0 > Domain-0 0 1 1 r-- 24.6 1With my setup, it''s like this # cat /proc/cpuinfo | grep processor processor : 0 # xm vcpu-list 0 Name ID VCPU CPU State Time(s) CPU Affinity Domain-0 0 0 0 r-- 93253.0 0 Domain-0 0 1 - --p 579.8 1 Domain-0 0 2 - --p 11.4 2 Domain-0 0 3 - --p 8.0 3 Domain-0 0 4 - --p 8.0 4 Domain-0 0 5 - --p 13.5 5 Domain-0 0 6 - --p 11.7 6 Domain-0 0 7 - --p 10.0 7 the benefit is that I can hotplug other cpus to dom0 if necessary later.> the point remains, however, that ntpd withOUT xen''s presence, on > any/all hardware I have, runs quiet & stable. clearly, "xen" is > involved. whether it''s the opensuse implementation of kernel-xen or > Xen, or something inherent in Xen''s design, is unknown atm ...Hope you can solve this problem. When you do, please share your results. -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 04:12 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
> Hope you can solve this problem. When you do, please share your results.@ novell say it''s xen & ntp @ ntp say it''s novell & xen @ xen say it''s novell & ntp this does not bode well :-( _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-15 04:20 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 11:12 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:>> Hope you can solve this problem. When you do, please share your results. > > @ novell say it''s xen & ntp > @ ntp say it''s novell & xen > @ xen say it''s novell & ntp > > this does not bode well :-(That''s why I suggested you use a workaround: use 2.6.18 kernel, either from RHEL or from xen.org :D If this works -> novell problem If this doesn''t work -> xen problem -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 04:43 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Thu, Jan 14, 2010 at 8:20 PM, Fajar A. Nugraha <fajar@fajar.net> wrote:> That''s why I suggested you use a workaround: use 2.6.18 kernel, either > from RHEL or from xen.org :D > If this works -> novell problem > If this doesn''t work -> xen problemi have no idea how to boot opensuse from a redhat kernel rpm -- or even install it (e.g. http://mirror.centos.org/centos/5.4/centosplus/x86_64/RPMS/kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm) without complaints about deps etc. or do you mean build a new from src @xen or @redhat?? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-15 07:14 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 11:43 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> On Thu, Jan 14, 2010 at 8:20 PM, Fajar A. Nugraha <fajar@fajar.net> wrote: >> That''s why I suggested you use a workaround: use 2.6.18 kernel, either >> from RHEL or from xen.org :D >> If this works -> novell problem >> If this doesn''t work -> xen problem > > i have no idea how to boot opensuse from a redhat kernel rpm -- or > even install it (e.g. > http://mirror.centos.org/centos/5.4/centosplus/x86_64/RPMS/kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm) > without complaints about deps etc. > > or do you mean build a new from src @xen or @redhat??xen.org''s 2.6.18 kernel -> build from src, http://www.xen.org/products/xen_source.html RHEL/Centos kernel rpm -> quite tricky, admittedly :D I suppose you could try something like this wget -c http://mirror.centos.org/centos/5.4/centosplus/x86_64/RPMS/kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm mkdir tmp-root cd tmp-root rpm2cpio ../kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm | cpio -idmv cp -a ./boot/* /boot cp -a ./lib/modules/* /lib/modules that is, extract the contents of kernel-xen rpm, and take only what''s in /boot and /lib/modules. After this you need to build initrd. The command should be mkinitrd (haven''t done this for a long time in opensuse, so I forgot the exact syntax). Then you need to edit to add the kernel and initrd entry to /boot/grub/menu.lst. -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 07:42 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
ok, both _should_ be simple enough. On Thu, Jan 14, 2010 at 11:14 PM, Fajar A. Nugraha <fajar@fajar.net> wrote:> xen.org''s 2.6.18 kernel -> build from src, > http://www.xen.org/products/xen_source.htmltried that; scads of errors @ make. tried a build of opensuse kernel src, just to check my procedure -- works fine.> RHEL/Centos kernel rpm -> quite tricky, admittedly :D > I suppose you could try something like this > > wget -c http://mirror.centos.org/centos/5.4/centosplus/x86_64/RPMS/kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm > mkdir tmp-root > cd tmp-root > rpm2cpio ../kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm | cpio -idmv > cp -a ./boot/* /boot > cp -a ./lib/modules/* /lib/modules > > that is, extract the contents of kernel-xen rpm, and take only what''s > in /boot and /lib/modules. After this you need to build initrd. The > command should be mkinitrd (haven''t done this for a long time in > opensuse, so I forgot the exact syntax). Then you need to edit to add > the kernel and initrd entry to /boot/grub/menu.lst.it *is* just mkinitrd ... but, when i do, mkinitrd Kernel image: /boot/vmlinuz-2.6.18-164.10.1.el5.centos.plusxen Initrd image: /boot/initrd-2.6.18-164.10.1.el5.centos.plusxen Root device: /dev/root (mounted on / as ext4) Resume device: /dev/md2 setup-md.sh: md1 found multiple times modprobe: Could not load /lib/modules/2.6.18-164.10.1.el5.centos.plusxen/modules.dep: No such file or directory WARNING: no dependencies for kernel module ''raid0'' found. modprobe: Could not load /lib/modules/2.6.18-164.10.1.el5.centos.plusxen/modules.dep: No such file or directory WARNING: no dependencies for kernel module ''raid1'' found. modprobe: Could not load /lib/modules/2.6.18-164.10.1.el5.centos.plusxen/modules.dep: No such file or directory WARNING: no dependencies for kernel module ''raid456'' found. ... ... ... where do i get the modules.dep file? afaict, it''s not in the redhat .rpm''s extract. thanks. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-15 08:10 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Thu, Jan 14, 2010 at 11:42:43PM -0800, mail ignored wrote:> ok, both _should_ be simple enough. > > On Thu, Jan 14, 2010 at 11:14 PM, Fajar A. Nugraha <fajar@fajar.net> wrote: > > xen.org''s 2.6.18 kernel -> build from src, > > http://www.xen.org/products/xen_source.html > > tried that; scads of errors @ make. > > tried a build of opensuse kernel src, just to check my procedure -- works fine. > > > RHEL/Centos kernel rpm -> quite tricky, admittedly :D > > I suppose you could try something like this > > > > wget -c http://mirror.centos.org/centos/5.4/centosplus/x86_64/RPMS/kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm > > mkdir tmp-root > > cd tmp-root > > rpm2cpio ../kernel-xen-2.6.18-164.10.1.el5.centos.plus.x86_64.rpm | cpio -idmv > > cp -a ./boot/* /boot > > cp -a ./lib/modules/* /lib/modules > > > > that is, extract the contents of kernel-xen rpm, and take only what''s > > in /boot and /lib/modules. After this you need to build initrd. The > > command should be mkinitrd (haven''t done this for a long time in > > opensuse, so I forgot the exact syntax). Then you need to edit to add > > the kernel and initrd entry to /boot/grub/menu.lst. > > it *is* just mkinitrd ... but, when i do, > > mkinitrd > > Kernel image: /boot/vmlinuz-2.6.18-164.10.1.el5.centos.plusxen > Initrd image: /boot/initrd-2.6.18-164.10.1.el5.centos.plusxen > Root device: /dev/root (mounted on / as ext4) > Resume device: /dev/md2 > setup-md.sh: md1 found multiple times > modprobe: Could not load > /lib/modules/2.6.18-164.10.1.el5.centos.plusxen/modules.dep: No such > file or directory > WARNING: no dependencies for kernel module ''raid0'' found. > modprobe: Could not load > /lib/modules/2.6.18-164.10.1.el5.centos.plusxen/modules.dep: No such > file or directory > WARNING: no dependencies for kernel module ''raid1'' found. > modprobe: Could not load > /lib/modules/2.6.18-164.10.1.el5.centos.plusxen/modules.dep: No such > file or directory > WARNING: no dependencies for kernel module ''raid456'' found. > ... > ... > ... > > where do i get the modules.dep file? afaict, it''s not in the redhat > .rpm''s extract. >Run "depmod -a 2.6.18-164.10.1.el5.centos.plusxen". -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 16:30 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 12:10 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> Run "depmod -a 2.6.18-164.10.1.el5.centos.plusxen".i had run ''depmod -a'', as i always do when building opensuse kernels, with no effect. i did not realize that "depmod -a 2.6.18-164.10.1.el5.centos.plusxen" -- which works! -- was required. thanks. sorry this is becoming a treatise on kernel-building, but i appreciate the necessity to try to rule out xen ... atm, i''ve extracted/installed from the centos rpm, as Fajar instructed. i''ve also run ''mkinitrd -f lvm2 -f md'' so now i have, ls -al /boot/*el* -rw-r--r-- 1 root root 68951 2010-01-08 14:54 /boot/config-2.6.18-164.10.1.el5.centos.plusxen -rw-r--r-- 1 root root 8199432 2010-01-15 07:51 /boot/initrd-2.6.18-164.10.1.el5.centos.plusxen -rw-r--r-- 1 root root 116122 2010-01-08 14:55 /boot/symvers-2.6.18-164.10.1.el5.centos.plusxen.gz -rw-r--r-- 1 root root 1196044 2010-01-08 14:54 /boot/System.map-2.6.18-164.10.1.el5.centos.plusxen -rw-r--r-- 1 root root 2031384 2010-01-08 14:54 /boot/vmlinuz-2.6.18-164.10.1.el5.centos.plusxen -rw-r--r-- 1 root root 413719 2010-01-08 13:41 /boot/xen.gz-2.6.18-164.10.1.el5.centos.plus -rwxr-xr-x 1 root root 965728 2010-01-08 15:01 /boot/xen-syms-2.6.18-164.10.1.el5.centos.plus i''ve setup grub, to use the {vmlinuz,initrd}-2.6.18-164.10.1.el5.centos.plusxen. at boot, i see the hypervisor load, and the passoff to the centos kernel. but it then gets stuck in a loop, endlessly repeating "volume group VG0 not found". fwiw, "/boot" is on a linux native RAID-1 part, and "/" is on a LV on a RAID-1. of course, works fine @ boot to opensuse kernel(s). what else do I need to get that centos kernel to boot my box? Or is the 2.6.18* too old for support of LVM, RAID, etc (i doubt that ...)? thanks. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 16:58 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 8:30 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> what else do I need to get that centos kernel to boot my box? Or is > the 2.6.18* too old for support of LVM, RAID, etc (i doubt that ...)?fwiw, creating a "monster initrd", with ''mkinitrd -A'', although successful, fails to change the behavior -- still VolGrp not found. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 17:28 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 8:58 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> On Fri, Jan 15, 2010 at 8:30 AM, mail ignored <0.bugs.only.0@gmail.com> wrote: >> what else do I need to get that centos kernel to boot my box? Or is >> the 2.6.18* too old for support of LVM, RAID, etc (i doubt that ...)?is the centos/redhat 2.6.18 kernel ext4-friendly? opensuse 11.2 installs as ext4 by default, and i chose to stick with it @ install. my fstab includes, LABEL=ROOT / ext4 relatime,acl,user_xattr,journal_checksum,barrier=1 1 1 since "/" is using extents, i can''t re-mount it as ext3 ... _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 17:33 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 9:28 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> is the centos/redhat 2.6.18 kernel ext4-friendly?this http://etbe.coker.com.au/2010/01/05/ext4-and-rhel5-centos5/ suggests that ext4 should not be a problem with this kernel ... _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-15 22:34 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Sat, Jan 16, 2010 at 12:33 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> On Fri, Jan 15, 2010 at 9:28 AM, mail ignored <0.bugs.only.0@gmail.com> wrote: >> is the centos/redhat 2.6.18 kernel ext4-friendly? > > this > > http://etbe.coker.com.au/2010/01/05/ext4-and-rhel5-centos5/ > > suggests that ext4 should not be a problem with this kernel ...Last time I check it only recognize ext4dev. I''ll try this next week. What does your partition setup looks like? Is it the default opensuse layout, or did you customize it? last time I check grub can''t support /boot on raid-1. Also, does the Centos kernel recognize the disks succesfully (i.e. did you see messages about sda/sdb during boot)? -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-15 23:10 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi, On Fri, Jan 15, 2010 at 2:34 PM, Fajar A. Nugraha <fajar@fajar.net> wrote:> Last time I check it only recognize ext4dev. I''ll try this next week.in addition to the reference above, this, https://bugzilla.redhat.com/show_bug.cgi?id=485315#c13 suggests it does. thanks.> What does your partition setup looks like?sda, 1TB sdb, 1TB /dev/md0 = 150 MB, ext3, RAID-1 (/dev/sda1 + /dev/sdb1) /dev/md1 = *, raw LVM, RAID-1 (/dev/sda2 + /dev/sdb2) LVMs, /dev/VG0/ROOT 50GB, ext4 /dev/VG0/SWAP 2 GB, swap /dev/VG0/HOME *, ext4> Is it the default opensuse layout, or did you customize it?It''s certainly not the default. But it is fully supported by opensuse, and readily configurable in its Partitioner> last time I check grub can''t support /boot on raid-1.Works just fine on OpenSuse. Using Centos kernel, the hypervisor & kernel both seem to found/recognized just fine ... cref e.g., http://www.linuxfoundation.org/collaborate/workgroups/linux-raid/preventing_against_a_failing_disk#Prepare_for_boot " ... On Intel-compatibe hardware, there are two common boot loaders, grub and lilo. Both grub and lilo can only boot off a raid1. They cannot boot off any other software raid device type ... "> Also, does the Centos kernel recognize the disks > succesfully (i.e. did you see messages about sda/sdb during boot)?yes, both are there. it''s the VG that''s it''s complaining is not being found. although the messages scroll by too quickly (i''ll find a serial cable around here ...), i do see something about "udevadm ... Connection Refused ...". _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-16 19:17 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Thu, Jan 14, 2010 at 8:20 PM, Fajar A. Nugraha <fajar@fajar.net> wrote:>> @ novell say it''s xen & ntp >> @ ntp say it''s novell & xen >> @ xen say it''s novell & ntp >>> That''s why I suggested you use a workaround: use 2.6.18 kernel, either > from RHEL or from xen.org :D > If this works -> novell problem > If this doesn''t work -> xen problemOn Sat, Jan 16, 2010 at 5:05 AM, Fajar A. Nugraha <fajar@fajar.net> wrote:> You might want to try latest pv_ops as well > (http://wiki.xensource.com/xenwiki/XenParavirtOps), which should be > easier to work with Opensuse since the version number is close.it seems it''s _not_ the kernel -- unless there''s a coincident kernel problem across source/vendor ... per http://lists.xensource.com/archives/html/xen-devel/2009-12/msg01127.html http://wiki.xensource.com/xenwiki/XenParavirtOps http://wiki.xensource.com/xenwiki/Kernel.org_Linux_on_Xen cd /home/workspace && rm -rf linux-2.6-xen git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git linux-2.6-xen cd linux-2.6-xen make clean cp -a .config .config-old wget -O .config http://pasik.reaktio.net/xen/pv_ops-dom0-debug/config-2.6.31.4-pvops_dom0-x86_64 make oldconfig ... Xen PCI Frontend (XEN_PCIDEV_FRONTEND) [N/m/y/?] (NEW) <-- m ... "IMQ" target support (NETFILTER_XT_TARGET_IMQ) [N/m/?] (NEW) <-- m ... IMQ (intermediate queueing device) support (IMQ) [M/y/?] (NEW) <-- [enter] ... Block-device tap backend driver (XEN_BLKDEV_TAP) [N/m/y/?] (NEW) <-- m ... make bzImage make modules make modules_install strip vmlinux -o vmlinux-stripped gzip vmlinux-stripped -c > vmlinuz-gzipped ls -al .config \ System.map \ arch/x86/boot/bzImage \ vmlinux \ vmlinux-stripped \ vmlinuz-gzipped cp -a .config /boot/config-2.6.31.6 cp -a System.map /boot/System.map-2.6.31.6 cp -a arch/x86/boot/bzImage /boot/vmlinuz-2.6.31.6 mkinitrd ls -al /boot/*2.6.31.6* -rw-r--r-- 1 root root 98510 2010-01-16 03:02 /boot/config-2.6.31.6 -rw-r--r-- 1 root root 21304611 2010-01-16 04:14 /boot/initrd-2.6.31.6 -rw-r--r-- 1 root root 1940879 2010-01-16 03:14 /boot/System.map-2.6.31.6 -rw-r--r-- 1 root root 3751936 2010-01-16 03:14 /boot/vmlinuz-2.6.31.6 vi /boot/grub/menu.lst ... title xen.git root (hd0,0) kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 module /vmlinuz-2.6.31.6 root=/dev/VG0/ROOT rootfstype=ext4 rootflags=journal_checksum resume=/dev/VG0/SWAP nomodeset selinux=0 SELINUX_INIT=NO vga=0x31a console=tty0 console=xvc0,57600 module /initrd-2.6.31.6 title xen.git DISABLED root (hd0,0) kernel /vmlinuz-2.6.31.6 root=/dev/VG_SVR/LV_SVR_DOM0_ROOT rootfstype=ext4 rootflags=journal_checksum resume=/dev/VG_SVR/LV_SVR_DOM0_SWAP nomodeset showopts splash=silent selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 console=ttyS0,57600n8 module /initrd-2.6.31.6 ... boot to "xen.git" uname -a Linux test 2.6.31.6 #1 SMP Sat Jan 16 03:14:19 PST 2010 x86_64 x86_64 x86_64 GNU/Linux lsb_release -a LSB Version: core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:graphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:graphics-4.0-noarch Distributor ID: SUSE LINUX Description: openSUSE 11.2 (x86_64) Release: 11.2 Codename: n/a service ntp stop rm /etc/adjtime /var/ rm /var/lib/ntp/drift/ntp.drift cat << EOF > /etc/sysconfig/clock HWCLOCK="--utc" SYSTOHC="yes" TIMEZONE="America/Los_Angeles" DEFAULT_TIMEZONE="US/Pacific" EOF hwclock --systohc --utc service ntp start sntp clock.isc.org && date && hwclock --utc --show 2010 Jan 16 04:37:09.000729 + 1.435060 +/- 0.028944 secs Sat Jan 16 04:37:11 PST 2010 Sat 16 Jan 2010 04:37:11 AM PST -0.345191 seconds ... (wait ~ 6 hours) ... date Sat Jan 16 11:01:42 PST 2010 ntpq -p -c rv assID=0 status=06c4 leap_none, sync_ntp, 12 events, event_peer/strat_chg, version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.6", leap=00, stratum=2, precision=-20, rootdelay=14.718, rootdispersion=135.995, peer=11390, refid=66.220.9.122, reftime=cefc89b2.a36f1342 Sat, Jan 16 2010 10:53:38.638, poll=6, clock=cefc8b82.7e523feb Sat, Jan 16 2010 11:01:22.493, state=2, offset=-115.130, frequency=-500.000, jitter=464.242, noise=29.343, stability=154.680 remote refid st t when poll reach delay offset jitter ============================================================================= *clock.fmt.he.ne .PPS. 1 u 29 64 377 21.438 -811.68 462.837 +otc2.psu.edu 129.6.15.29 2 u 7 64 377 109.399 -850.99 470.522 +clock.isc.org .GPS. 1 u 7 64 377 21.440 -854.82 468.274 +clock.sjc.he.ne .CDMA. 1 u 49 64 377 16.889 -780.55 469.481 +zorro.sf-bay.or 216.218.254.202 2 u 59 64 377 20.366 -762.14 458.947 +nist1.aol-ca.tr .ACTS. 1 u 38 64 377 16.260 -795.87 463.170 boot to "xen.git DISABLED" uname -a Linux test 2.6.31.6 #1 SMP Sat Jan 16 09:14:19 PST 2010 x86_64 x86_64 x86_64 GNU/Linux service ntp stop rm /etc/adjtime /var/ rm /var/lib/ntp/drift/ntp.drift service ntp start ... (wait ~ 10 minutes) ... date Sat Jan 16 11:15:05 PST 2010 ntpq -p -c rv assID=0 status=0664 leap_none, sync_ntp, 6 events, event_peer/strat_chg, version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.6", leap=00, stratum=2, precision=-20, rootdelay=22.750, rootdispersion=975.184, peer=43091, refid=66.220.9.122, reftime=cefc8c7c.35ca466a Sat, Jan 16 2010 11:05:32.210, poll=6, clock=cefc8eb9.25103a1f Sat, Jan 16 2010 11:15:05.144, state=3, offset=27.534, frequency=0.000, jitter=6.740, noise=16.331, stability=0.000 remote refid st t when poll reach delay offset jitter ============================================================================= *clock.fmt.he.ne .PPS. 1 u 4 64 377 18.112 10.541 5.892 +otc2.psu.edu 147.84.59.145 2 u 64 64 377 101.103 12.106 8.621 +clock.isc.org .GPS. 1 u 57 64 377 14.990 4.830 9.622 +clock.sjc.he.ne .CDMA. 1 u 59 64 377 16.503 9.299 9.440 +zorro.sf-bay.or 216.218.254.202 2 u 56 64 377 22.212 6.323 11.031 +nist1.aol-ca.tr .ACTS. 1 u 60 64 377 17.076 14.316 9.260 in summary, kernel OS/distro xen timekeeping ------------------------------------- --------------- ------- ------------ opensuse kernel-default-2.6.31.8-0.1 opensuse 11.2 (n/a) OK opensuse kernel-default-2.6.32-41 opensuse 11.2 (n/a) OK opensuse kernel-xen-2.6.31.8-0.1 opensuse 11.2 3.4.x NOISY/FAIL opensuse kernel-xen-2.6.32-41 opensuse 11.2 3.4.x NOISY/FAIL opensuse kernel-xen-2.6.31.8-0.1 opensuse 11.2 4.0.x NOISY/FAIL opensuse kernel-xen-2.6.32-41 opensuse 11.2 4.0.x NOISY/FAIL xen.git kernel-2.6.31.6 (pvops Dom0) opensuse 11.2 4.0.x NOISY/FAIL xen.git kernel-2.6.31.6 opensuse 11.2 (n/a) OK i.e., 3 different kernel versions, from two different vendors, have low-noise/accurate timekeeping in NON-xen configs. boot any of those kernels with either v34x or v40x hypervisor in place, and timekeeping suffers from unusably excessive jitter & offset. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-17 20:03 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Fri, Jan 15, 2010 at 09:33:20AM -0800, mail ignored wrote:> On Fri, Jan 15, 2010 at 9:28 AM, mail ignored <0.bugs.only.0@gmail.com> wrote: > > is the centos/redhat 2.6.18 kernel ext4-friendly? > > this > > http://etbe.coker.com.au/2010/01/05/ext4-and-rhel5-centos5/ > > suggests that ext4 should not be a problem with this kernel ...RHEL/CentOS 5 -164 kernel does support ext4. Please pastebin the full console log during system start up so we can see what happens. -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-17 20:22 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
Keir and others, Please check the table in the end of the mail below.. is there hypervisor (boot time) option to make it behave better for ntpd? -- Pasi On Sat, Jan 16, 2010 at 11:17:36AM -0800, mail ignored wrote:> On Thu, Jan 14, 2010 at 8:20 PM, Fajar A. Nugraha <fajar@fajar.net> wrote: > >> @ novell say it''s xen & ntp > >> @ ntp say it''s novell & xen > >> @ xen say it''s novell & ntp > >> > > > That''s why I suggested you use a workaround: use 2.6.18 kernel, either > > from RHEL or from xen.org :D > > If this works -> novell problem > > If this doesn''t work -> xen problem > > On Sat, Jan 16, 2010 at 5:05 AM, Fajar A. Nugraha <fajar@fajar.net> wrote: > > You might want to try latest pv_ops as well > > (http://wiki.xensource.com/xenwiki/XenParavirtOps), which should be > > easier to work with Opensuse since the version number is close. > > it seems it''s _not_ the kernel -- unless there''s a coincident kernel > problem across source/vendor ... > > per > http://lists.xensource.com/archives/html/xen-devel/2009-12/msg01127.html > http://wiki.xensource.com/xenwiki/XenParavirtOps > http://wiki.xensource.com/xenwiki/Kernel.org_Linux_on_Xen > > cd /home/workspace && rm -rf linux-2.6-xen > git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git > linux-2.6-xen > cd linux-2.6-xen > make clean > cp -a .config .config-old > wget -O .config > http://pasik.reaktio.net/xen/pv_ops-dom0-debug/config-2.6.31.4-pvops_dom0-x86_64 > make oldconfig > ... > Xen PCI Frontend (XEN_PCIDEV_FRONTEND) [N/m/y/?] (NEW) <-- m > ... > "IMQ" target support (NETFILTER_XT_TARGET_IMQ) [N/m/?] (NEW) <-- m > ... > IMQ (intermediate queueing device) support (IMQ) [M/y/?] (NEW) <-- [enter] > ... > Block-device tap backend driver (XEN_BLKDEV_TAP) [N/m/y/?] (NEW) <-- m > ... > > make bzImage > make modules > make modules_install > strip vmlinux -o vmlinux-stripped > gzip vmlinux-stripped -c > vmlinuz-gzipped > > ls -al .config \ > System.map \ > arch/x86/boot/bzImage \ > vmlinux \ > vmlinux-stripped \ > vmlinuz-gzipped > > cp -a .config /boot/config-2.6.31.6 > cp -a System.map /boot/System.map-2.6.31.6 > cp -a arch/x86/boot/bzImage /boot/vmlinuz-2.6.31.6 > mkinitrd > ls -al /boot/*2.6.31.6* > -rw-r--r-- 1 root root 98510 2010-01-16 03:02 /boot/config-2.6.31.6 > -rw-r--r-- 1 root root 21304611 2010-01-16 04:14 /boot/initrd-2.6.31.6 > -rw-r--r-- 1 root root 1940879 2010-01-16 03:14 /boot/System.map-2.6.31.6 > -rw-r--r-- 1 root root 3751936 2010-01-16 03:14 /boot/vmlinuz-2.6.31.6 > > > vi /boot/grub/menu.lst > ... > title xen.git > root (hd0,0) > kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M > vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 > module /vmlinuz-2.6.31.6 root=/dev/VG0/ROOT rootfstype=ext4 > rootflags=journal_checksum resume=/dev/VG0/SWAP nomodeset selinux=0 > SELINUX_INIT=NO vga=0x31a console=tty0 console=xvc0,57600 > module /initrd-2.6.31.6 > > title xen.git DISABLED > root (hd0,0) > kernel /vmlinuz-2.6.31.6 root=/dev/VG_SVR/LV_SVR_DOM0_ROOT > rootfstype=ext4 rootflags=journal_checksum > resume=/dev/VG_SVR/LV_SVR_DOM0_SWAP nomodeset showopts splash=silent > selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 > console=ttyS0,57600n8 > module /initrd-2.6.31.6 > ... > > > boot to "xen.git" > > uname -a > Linux test 2.6.31.6 #1 SMP Sat Jan 16 03:14:19 PST 2010 x86_64 > x86_64 x86_64 GNU/Linux > lsb_release -a > LSB Version: > core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:graphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:graphics-4.0-noarch > Distributor ID: SUSE LINUX > Description: openSUSE 11.2 (x86_64) > Release: 11.2 > Codename: n/a > > service ntp stop > rm /etc/adjtime /var/ > rm /var/lib/ntp/drift/ntp.drift > > cat << EOF > /etc/sysconfig/clock > HWCLOCK="--utc" > SYSTOHC="yes" > TIMEZONE="America/Los_Angeles" > DEFAULT_TIMEZONE="US/Pacific" > EOF > hwclock --systohc --utc > service ntp start > > sntp clock.isc.org && date && hwclock --utc --show > 2010 Jan 16 04:37:09.000729 + 1.435060 +/- 0.028944 secs > Sat Jan 16 04:37:11 PST 2010 > Sat 16 Jan 2010 04:37:11 AM PST -0.345191 seconds > > ... (wait ~ 6 hours) ... > > date > Sat Jan 16 11:01:42 PST 2010 > ntpq -p -c rv > assID=0 status=06c4 leap_none, sync_ntp, 12 events, event_peer/strat_chg, > version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", > processor="x86_64", system="Linux/2.6.31.6", leap=00, stratum=2, > precision=-20, rootdelay=14.718, rootdispersion=135.995, peer=11390, > refid=66.220.9.122, > reftime=cefc89b2.a36f1342 Sat, Jan 16 2010 10:53:38.638, poll=6, > clock=cefc8b82.7e523feb Sat, Jan 16 2010 11:01:22.493, state=2, > offset=-115.130, frequency=-500.000, jitter=464.242, noise=29.343, > stability=154.680 > remote refid st t when poll reach delay offset jitter > =============================================================================> *clock.fmt.he.ne .PPS. 1 u 29 64 377 21.438 -811.68 462.837 > +otc2.psu.edu 129.6.15.29 2 u 7 64 377 109.399 -850.99 470.522 > +clock.isc.org .GPS. 1 u 7 64 377 21.440 -854.82 468.274 > +clock.sjc.he.ne .CDMA. 1 u 49 64 377 16.889 -780.55 469.481 > +zorro.sf-bay.or 216.218.254.202 2 u 59 64 377 20.366 -762.14 458.947 > +nist1.aol-ca.tr .ACTS. 1 u 38 64 377 16.260 -795.87 463.170 > > boot to "xen.git DISABLED" > > uname -a > Linux test 2.6.31.6 #1 SMP Sat Jan 16 09:14:19 PST 2010 x86_64 x86_64 > x86_64 GNU/Linux > > service ntp stop > rm /etc/adjtime /var/ > rm /var/lib/ntp/drift/ntp.drift > service ntp start > > ... (wait ~ 10 minutes) ... > date > Sat Jan 16 11:15:05 PST 2010 > ntpq -p -c rv > assID=0 status=0664 leap_none, sync_ntp, 6 events, event_peer/strat_chg, > version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", > processor="x86_64", system="Linux/2.6.31.6", leap=00, stratum=2, > precision=-20, rootdelay=22.750, rootdispersion=975.184, peer=43091, > refid=66.220.9.122, > reftime=cefc8c7c.35ca466a Sat, Jan 16 2010 11:05:32.210, poll=6, > clock=cefc8eb9.25103a1f Sat, Jan 16 2010 11:15:05.144, state=3, > offset=27.534, frequency=0.000, jitter=6.740, noise=16.331, > stability=0.000 > remote refid st t when poll reach delay offset jitter > =============================================================================> *clock.fmt.he.ne .PPS. 1 u 4 64 377 18.112 10.541 5.892 > +otc2.psu.edu 147.84.59.145 2 u 64 64 377 101.103 12.106 8.621 > +clock.isc.org .GPS. 1 u 57 64 377 14.990 4.830 9.622 > +clock.sjc.he.ne .CDMA. 1 u 59 64 377 16.503 9.299 9.440 > +zorro.sf-bay.or 216.218.254.202 2 u 56 64 377 22.212 6.323 11.031 > +nist1.aol-ca.tr .ACTS. 1 u 60 64 377 17.076 14.316 9.260 > > in summary, > > kernel OS/distro xen timekeeping > ------------------------------------- --------------- ------- ------------ > opensuse kernel-default-2.6.31.8-0.1 opensuse 11.2 (n/a) OK > opensuse kernel-default-2.6.32-41 opensuse 11.2 (n/a) OK > opensuse kernel-xen-2.6.31.8-0.1 opensuse 11.2 3.4.x NOISY/FAIL > opensuse kernel-xen-2.6.32-41 opensuse 11.2 3.4.x NOISY/FAIL > opensuse kernel-xen-2.6.31.8-0.1 opensuse 11.2 4.0.x NOISY/FAIL > opensuse kernel-xen-2.6.32-41 opensuse 11.2 4.0.x NOISY/FAIL > xen.git kernel-2.6.31.6 (pvops Dom0) opensuse 11.2 4.0.x NOISY/FAIL > xen.git kernel-2.6.31.6 opensuse 11.2 (n/a) OK > > i.e., > > 3 different kernel versions, from two different vendors, have > low-noise/accurate timekeeping in NON-xen configs. > > boot any of those kernels with either v34x or v40x hypervisor in > place, and timekeeping suffers from unusably excessive jitter & > offset. > > _______________________________________________ > Xen-users mailing list > Xen-users@lists.xensource.com > http://lists.xensource.com/xen-users_______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-17 20:23 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Sun, Jan 17, 2010 at 12:03 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> RHEL/CentOS 5 -164 kernel does support ext4.Thanks for verifying.> Please pastebin the full console log during system start up so we can see what happens.Is continuing to work with/on getting RHEL/CentOS 5 kernel-xen to boot of any particular use/value to _this_ issue, given the successful build/boot on xen.git''s pvops kernel (results posted above)? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-17 20:25 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Sun, Jan 17, 2010 at 12:22 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> Keir and others, > > Please check the table in the end of the mail below.. is there > hypervisor (boot time) option to make it behave better for ntpd?Thanks for forwarding this ... _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Keir Fraser
2010-Jan-17 21:33 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
What platform timer does Xen say is initialised during boot? Possibly Xen''s ''clocksource='' boot option could be used to select a different platform timer exhibiting less jitter. Tbh I don''t think we''ve ever done a detailed analysis of ntp performance on Xen before -- this is something we might be able to improve. -- Keir On 17/01/2010 20:22, "Pasi Kärkkäinen" <pasik@iki.fi> wrote:> > Keir and others, > > Please check the table in the end of the mail below.. is there > hypervisor (boot time) option to make it behave better for ntpd? > > -- Pasi > > On Sat, Jan 16, 2010 at 11:17:36AM -0800, mail ignored wrote: >> On Thu, Jan 14, 2010 at 8:20 PM, Fajar A. Nugraha <fajar@fajar.net> wrote: >>>> @ novell say it''s xen & ntp >>>> @ ntp say it''s novell & xen >>>> @ xen say it''s novell & ntp >>>> >> >>> That''s why I suggested you use a workaround: use 2.6.18 kernel, either >>> from RHEL or from xen.org :D >>> If this works -> novell problem >>> If this doesn''t work -> xen problem >> >> On Sat, Jan 16, 2010 at 5:05 AM, Fajar A. Nugraha <fajar@fajar.net> wrote: >>> You might want to try latest pv_ops as well >>> (http://wiki.xensource.com/xenwiki/XenParavirtOps), which should be >>> easier to work with Opensuse since the version number is close. >> >> it seems it''s _not_ the kernel -- unless there''s a coincident kernel >> problem across source/vendor ... >> >> per >> http://lists.xensource.com/archives/html/xen-devel/2009-12/msg01127.html >> http://wiki.xensource.com/xenwiki/XenParavirtOps >> http://wiki.xensource.com/xenwiki/Kernel.org_Linux_on_Xen >> >> cd /home/workspace && rm -rf linux-2.6-xen >> git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git >> linux-2.6-xen >> cd linux-2.6-xen >> make clean >> cp -a .config .config-old >> wget -O .config >> http://pasik.reaktio.net/xen/pv_ops-dom0-debug/config-2.6.31.4-pvops_dom0-x86 >> _64 >> make oldconfig >> ... >> Xen PCI Frontend (XEN_PCIDEV_FRONTEND) [N/m/y/?] (NEW) <-- m >> ... >> "IMQ" target support (NETFILTER_XT_TARGET_IMQ) [N/m/?] (NEW) <-- m >> ... >> IMQ (intermediate queueing device) support (IMQ) [M/y/?] (NEW) <-- >> [enter] >> ... >> Block-device tap backend driver (XEN_BLKDEV_TAP) [N/m/y/?] (NEW) <-- m >> ... >> >> make bzImage >> make modules >> make modules_install >> strip vmlinux -o vmlinux-stripped >> gzip vmlinux-stripped -c > vmlinuz-gzipped >> >> ls -al .config \ >> System.map \ >> arch/x86/boot/bzImage \ >> vmlinux \ >> vmlinux-stripped \ >> vmlinuz-gzipped >> >> cp -a .config /boot/config-2.6.31.6 >> cp -a System.map /boot/System.map-2.6.31.6 >> cp -a arch/x86/boot/bzImage /boot/vmlinuz-2.6.31.6 >> mkinitrd >> ls -al /boot/*2.6.31.6* >> -rw-r--r-- 1 root root 98510 2010-01-16 03:02 /boot/config-2.6.31.6 >> -rw-r--r-- 1 root root 21304611 2010-01-16 04:14 /boot/initrd-2.6.31.6 >> -rw-r--r-- 1 root root 1940879 2010-01-16 03:14 /boot/System.map-2.6.31.6 >> -rw-r--r-- 1 root root 3751936 2010-01-16 03:14 /boot/vmlinuz-2.6.31.6 >> >> >> vi /boot/grub/menu.lst >> ... >> title xen.git >> root (hd0,0) >> kernel /xen.gz loglvl=all loglvl_guest=all dom0_mem=1024M >> vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 >> module /vmlinuz-2.6.31.6 root=/dev/VG0/ROOT rootfstype=ext4 >> rootflags=journal_checksum resume=/dev/VG0/SWAP nomodeset selinux=0 >> SELINUX_INIT=NO vga=0x31a console=tty0 console=xvc0,57600 >> module /initrd-2.6.31.6 >> >> title xen.git DISABLED >> root (hd0,0) >> kernel /vmlinuz-2.6.31.6 root=/dev/VG_SVR/LV_SVR_DOM0_ROOT >> rootfstype=ext4 rootflags=journal_checksum >> resume=/dev/VG_SVR/LV_SVR_DOM0_SWAP nomodeset showopts splash=silent >> selinux=0 SELINUX_INIT=NO elevator=cfq vga=0x31a console=tty0 >> console=ttyS0,57600n8 >> module /initrd-2.6.31.6 >> ... >> >> >> boot to "xen.git" >> >> uname -a >> Linux test 2.6.31.6 #1 SMP Sat Jan 16 03:14:19 PST 2010 x86_64 >> x86_64 x86_64 GNU/Linux >> lsb_release -a >> LSB Version: >> core-2.0-noarch:core-3.2-noarch:core-4.0-noarch:core-2.0-x86_64:core-3.2-x86_ >> 64:core-4.0-x86_64:desktop-4.0-amd64:desktop-4.0-noarch:graphics-2.0-amd64:gr >> aphics-2.0-noarch:graphics-3.2-amd64:graphics-3.2-noarch:graphics-4.0-amd64:g >> raphics-4.0-noarch >> Distributor ID: SUSE LINUX >> Description: openSUSE 11.2 (x86_64) >> Release: 11.2 >> Codename: n/a >> >> service ntp stop >> rm /etc/adjtime /var/ >> rm /var/lib/ntp/drift/ntp.drift >> >> cat << EOF > /etc/sysconfig/clock >> HWCLOCK="--utc" >> SYSTOHC="yes" >> TIMEZONE="America/Los_Angeles" >> DEFAULT_TIMEZONE="US/Pacific" >> EOF >> hwclock --systohc --utc >> service ntp start >> >> sntp clock.isc.org && date && hwclock --utc --show >> 2010 Jan 16 04:37:09.000729 + 1.435060 +/- 0.028944 secs >> Sat Jan 16 04:37:11 PST 2010 >> Sat 16 Jan 2010 04:37:11 AM PST -0.345191 seconds >> >> ... (wait ~ 6 hours) ... >> >> date >> Sat Jan 16 11:01:42 PST 2010 >> ntpq -p -c rv >> assID=0 status=06c4 leap_none, sync_ntp, 12 events, event_peer/strat_chg, >> version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", >> processor="x86_64", system="Linux/2.6.31.6", leap=00, stratum=2, >> precision=-20, rootdelay=14.718, rootdispersion=135.995, peer=11390, >> refid=66.220.9.122, >> reftime=cefc89b2.a36f1342 Sat, Jan 16 2010 10:53:38.638, poll=6, >> clock=cefc8b82.7e523feb Sat, Jan 16 2010 11:01:22.493, state=2, >> offset=-115.130, frequency=-500.000, jitter=464.242, noise=29.343, >> stability=154.680 >> remote refid st t when poll reach delay offset jitter >>=============================================================================>>>> *clock.fmt.he.ne .PPS. 1 u 29 64 377 21.438 -811.68 >> 462.837 >> +otc2.psu.edu 129.6.15.29 2 u 7 64 377 109.399 -850.99 >> 470.522 >> +clock.isc.org .GPS. 1 u 7 64 377 21.440 -854.82 >> 468.274 >> +clock.sjc.he.ne .CDMA. 1 u 49 64 377 16.889 -780.55 >> 469.481 >> +zorro.sf-bay.or 216.218.254.202 2 u 59 64 377 20.366 -762.14 >> 458.947 >> +nist1.aol-ca.tr .ACTS. 1 u 38 64 377 16.260 -795.87 >> 463.170 >> >> boot to "xen.git DISABLED" >> >> uname -a >> Linux test 2.6.31.6 #1 SMP Sat Jan 16 09:14:19 PST 2010 x86_64 x86_64 >> x86_64 GNU/Linux >> >> service ntp stop >> rm /etc/adjtime /var/ >> rm /var/lib/ntp/drift/ntp.drift >> service ntp start >> >> ... (wait ~ 10 minutes) ... >> date >> Sat Jan 16 11:15:05 PST 2010 >> ntpq -p -c rv >> assID=0 status=0664 leap_none, sync_ntp, 6 events, event_peer/strat_chg, >> version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", >> processor="x86_64", system="Linux/2.6.31.6", leap=00, stratum=2, >> precision=-20, rootdelay=22.750, rootdispersion=975.184, peer=43091, >> refid=66.220.9.122, >> reftime=cefc8c7c.35ca466a Sat, Jan 16 2010 11:05:32.210, poll=6, >> clock=cefc8eb9.25103a1f Sat, Jan 16 2010 11:15:05.144, state=3, >> offset=27.534, frequency=0.000, jitter=6.740, noise=16.331, >> stability=0.000 >> remote refid st t when poll reach delay offset jitter >>=============================================================================>>>> *clock.fmt.he.ne .PPS. 1 u 4 64 377 18.112 10.541 >> 5.892 >> +otc2.psu.edu 147.84.59.145 2 u 64 64 377 101.103 12.106 >> 8.621 >> +clock.isc.org .GPS. 1 u 57 64 377 14.990 4.830 >> 9.622 >> +clock.sjc.he.ne .CDMA. 1 u 59 64 377 16.503 9.299 >> 9.440 >> +zorro.sf-bay.or 216.218.254.202 2 u 56 64 377 22.212 6.323 >> 11.031 >> +nist1.aol-ca.tr .ACTS. 1 u 60 64 377 17.076 14.316 >> 9.260 >> >> in summary, >> >> kernel OS/distro xen timekeeping >> ------------------------------------- --------------- ------- ------------ >> opensuse kernel-default-2.6.31.8-0.1 opensuse 11.2 (n/a) OK >> opensuse kernel-default-2.6.32-41 opensuse 11.2 (n/a) OK >> opensuse kernel-xen-2.6.31.8-0.1 opensuse 11.2 3.4.x NOISY/FAIL >> opensuse kernel-xen-2.6.32-41 opensuse 11.2 3.4.x NOISY/FAIL >> opensuse kernel-xen-2.6.31.8-0.1 opensuse 11.2 4.0.x NOISY/FAIL >> opensuse kernel-xen-2.6.32-41 opensuse 11.2 4.0.x NOISY/FAIL >> xen.git kernel-2.6.31.6 (pvops Dom0) opensuse 11.2 4.0.x NOISY/FAIL >> xen.git kernel-2.6.31.6 opensuse 11.2 (n/a) OK >> >> i.e., >> >> 3 different kernel versions, from two different vendors, have >> low-noise/accurate timekeeping in NON-xen configs. >> >> boot any of those kernels with either v34x or v40x hypervisor in >> place, and timekeeping suffers from unusably excessive jitter & >> offset. >> >> _______________________________________________ >> Xen-users mailing list >> Xen-users@lists.xensource.com >> http://lists.xensource.com/xen-users_______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-17 22:21 UTC
[Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi, (should I keep replying to both lists now?) On Sun, Jan 17, 2010 at 1:33 PM, Keir Fraser <keir.fraser@eu.citrix.com> wrote:> What platform timer does Xen say is initialised during boot?xm dmesg | egrep -i "timer|clock" (XEN) ACPI: PM-Timer IO Port: 0x808 (XEN) ..TIMER: vector=0xF0 apic1=0 pin1=2 apic2=-1 pin2=-1 (XEN) Platform timer is 14.284MHz HPET (XEN) HPET: 4 timers in total, 3 timers will be used for broadcast (XEN) mcheck_poll: Machine check polling timer started.> Possibly Xen''s > ''clocksource='' boot option could be used to select a different platform > timer exhibiting less jitter.currently, i have " ... clocksource=xen ... " on these platforms, @ non-xen, uname -a Linux test 2.6.31.8-0.1-desktop #1 SMP PREEMPT 2009-12-15 23:55:40 +0100 x86_64 x86_64 x86_64 GNU/Linux cat /sys/devices/system/clocksource/clocksource0/available_clocksource tsc hpet acpi_pm cat /sys/devices/system/clocksource/clocksource0/current_clocksource tsc i''ve read "tsc" is a "good thing ..." ... but, @ xen, i''ve only, uname -a Linux test 2.6.31.8-0.1-xen #1 SMP 2009-12-15 23:55:40 +0100 x86_64 x86_64 x86_64 GNU/Linux cat /sys/devices/system/clocksource/clocksource0/available_clocksource xen jiffies cat /sys/devices/system/clocksource/clocksource0/current_clocksource xen i''ve already tried the apparently only other option, ''jiffies'', as a xen clocksource, echo "jiffies" > /sys/devices/system/clocksource/clocksource0/current_clocksource cat /sys/devices/system/clocksource/clocksource0/current_clocksource jiffies service ntp restart (also tried this experiemtn booting to clocksource=jiffies in grub conf -- no difference) but i''ve noticed no improvement in jitter/offset over time, assID=0 status=c654 sync_alarm, sync_ntp, 5 events, event_peer/strat_chg, version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.8-0.1-xen", leap=11, stratum=2, precision=-8, rootdelay=0.000, rootdispersion=10.755, peer=34524, refid=Qq, reftime=00000000.00000000 Wed, Feb 6 2036 22:28:16.000, poll=6, clock=cefe0ad1.036efb16 Sun, Jan 17 2010 14:18:49.013, state=2, offset=0.000, frequency=-362.896, jitter=568.398, noise=3.906, stability=0.000 remote refid st t when poll reach delay offset jitter =============================================================================+clock.fmt.he.ne .PPS. 1 u 26 64 377 18.351 -1363.2 572.904 +otc2.psu.edu 128.118.2.33 2 u 58 64 377 99.412 -1298.6 575.668 +clock.isc.org .GPS. 1 u 61 64 377 19.111 -1296.8 562.546 +clock.sjc.he.ne .CDMA. 1 u 35 64 377 24.181 -1341.8 561.600 +zorro.sf-bay.or 216.218.254.202 2 u 63 64 377 20.503 -1290.7 565.956 *nist1.aol-ca.tr .ACTS. 1 u 35 64 377 24.571 -1339.8 567.403 and, in both these results, & at initially startup, assID=0 status=c035 sync_alarm, sync_unspec, 3 events, event_clock_reset, version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.8-0.1-xen", leap=11, stratum=16, precision=-8, rootdelay=0.000, rootdispersion=0.930, peer=0, refid=STEP, reftime=00000000.00000000 Wed, Feb 6 2036 22:28:16.000, poll=4, clock=cefe0842.32f14db2 Sun, Jan 17 2010 13:26:54.198, state=4, offset=0.000, frequency=-362.896, jitter=5.498, noise=3.906, stability=0.000 remote refid st t when poll reach delay offset jitter ============================================================================= clock.fmt.he.ne .PPS. 1 u 17 64 1 20.834 -88.894 3.906 otc2.psu.edu 130.207.244.240 2 u 54 64 1 123.655 -12.796 3.906 clock.isc.org .GPS. 1 u 49 64 1 19.517 -34.400 3.906 clock.sjc.he.ne .CDMA. 1 u 25 64 1 16.966 -76.911 3.906 zorro.sf-bay.or 216.218.254.202 2 u 48 64 1 35.391 -29.823 3.906 nist1.aol-ca.tr .ACTS. 1 u 24 64 1 24.211 -74.151 3.906 -- i see "3.906 weirdness". notice the 3.906''s in the jitter (changes over time) & noise/offset (fixed over time) fields? looking around on the net, i''ve found no particular explanation for this; tbh, it doesn''t exactly breed confidence.> Tbh I don''t think we''ve ever done a detailed analysis of ntp performance on Xen before > -- this is something we might be able to improve.any help would be appreciated. thanks. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
mail ignored
2010-Jan-17 23:53 UTC
[Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Sun, Jan 17, 2010 at 1:33 PM, Keir Fraser <keir.fraser@eu.citrix.com> wrote:> Tbh I don''t think we''ve ever done a detailed > analysis of ntp performance on Xen before -- this is something we might be > able to improve.just fyi -- this may be all old info for you -- this "Kernel Model for Precision Timekeeping" http://www.eecis.udel.edu/~mills/ntp/html/kern.html and the site/info around it, seem to be the authoritative resource on timekeeping with ntp (function, debugging, etc). it''s been a great resource in understanding the expected behavior of ntp. i also found good general guidance in "Network Time Protocol (NTP) Overview and Configuration: An Internet2 Cookbook" http://e2epi.internet2.edu/npw/binder-docs/ntp-cookbook.pdf although it does include a # of quotes/references to Mills'' bias against Linux, e.g., “I have confronted Linux many, many times with broken this and that and concluded I will not attempt to examine, fix, or judge anything Linux. You are on your own.” —David L. Mills, 2004-10-20 so it''s not clear, at least to me, if there are in fact ''deeper'' issues with ntp at hand. hopefully that''s gotten better over the years ... hth. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Dan Magenheimer
2010-Jan-18 02:23 UTC
RE: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
I''m away from the office right now so can''t easily look this up but the following patch was submitted by Tim Deegan in March 2009 and may be relevant. Likely it is in Xen''s dom0 but not in the dom0 of the distro you are using. Apologies if I''ve missed an important point as I''m only skimming email today and just happened to have Tim''s March email in a folder. Re: [Xen-devel] [PATCH] linux-2.6.18-xen.hg: fix interaction of NTP and dom0->xen time updates> -----Original Message----- > From: mail ignored [mailto:0.bugs.only.0@gmail.com] > Sent: Sunday, January 17, 2010 4:53 PM > To: Keir Fraser > Cc: Fajar A. Nugraha; xen-devel@lists.xensource.com; Xen User-List > Subject: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits > extremely high jitter/noise? runs stable/quiet under non-xen kernel. > > > On Sun, Jan 17, 2010 at 1:33 PM, Keir Fraser > <keir.fraser@eu.citrix.com> wrote: > > Tbh I don''t think we''ve ever done a detailed > > analysis of ntp performance on Xen before -- this is > something we might be > > able to improve. > > just fyi -- this may be all old info for you -- this > > "Kernel Model for Precision Timekeeping" > http://www.eecis.udel.edu/~mills/ntp/html/kern.html > > and the site/info around it, seem to be the authoritative resource on > timekeeping with ntp (function, debugging, etc). > > it''s been a great resource in understanding the expected > behavior of ntp. > > i also found good general guidance in > > "Network Time Protocol (NTP) Overview and Configuration: An > Internet2 Cookbook" > http://e2epi.internet2.edu/npw/binder-docs/ntp-cookbook.pdf > > although it does include a # of quotes/references to Mills'' bias > against Linux, e.g., > > “I have confronted Linux many, many times with broken this and that > and concluded I will not attempt to examine, fix, or judge anything > Linux. You are on your own.” —David L. Mills, 2004-10-20 > > so it''s not clear, at least to me, if there are in fact ''deeper'' > issues with ntp at hand. hopefully that''s gotten better over the > years ... > > hth. > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel >_______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-18 03:00 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Sun, Jan 17, 2010 at 6:23 PM, Dan Magenheimer <dan.magenheimer@oracle.com> wrote:> I''m away from the office right now so can''t easily look this up > but the following patch was submitted by Tim Deegan in March 2009 > and may be relevant. Likely it is in Xen''s dom0 but not in > the dom0 of the distro you are using. > > Apologies if I''ve missed an important point as I''m only > skimming email today and just happened to have Tim''s > March email in a folder.reading @ the patch thread, http://tinyurl.com/yjnj6wa for reference ... checking in opensuse''s src trees, cd /usr/src find . | grep x86 | grep time | grep xen | grep "\.c"$ ./arch/x86/kernel/time_32-xen.c ./arch/x86/xen/time.c time_32-xen.c is posted -> http://pastebin.com/f72c6de07 time.c -> http://pastebin.com/f3b4a7fdd checking in xen.git''s pvops tree, find . | grep x86 | grep time | grep xen | grep "\.c"$ ./arch/x86/xen/time.c -> http://pastebin.com/f3409ea7a. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Fajar A. Nugraha
2010-Jan-18 03:47 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 3:23 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:> Is continuing to work with/on getting RHEL/CentOS 5 kernel-xen to boot > of any particular use/value to _this_ issue, given the successful > build/boot on xen.git''s pvops kernel (results posted above)?It''s basically like this: The "stable", officially supported kernel version for xen as of now is still 2.6.18. When you encounter possible kernel-related problems, it''s usually best to check if the problem is present when using that kernel version. If it works, then you have one quick workaorund: simply use 2.6.18 kernel. I recommended RHEL/Centos5 kernel since it''s well maintained in terms of security fixes. pv_ops and suse''s xen kernel is based on newer kernel version. pv_ops will be the next official xen kernel, but currently it still have some bugs on it. For me it was the combination of suspend/resume and cpu hotplug. There were also problems with iperf (search list archive). Some of the bugs may take a long time to fix. So now the question becomes how quickly you need to solve this issue. If you can live with it, then I suggest focus all efforts on fixing pv_ops kernel (since you said that novell guys are unable to find the problem). If you need it to work quickly, then you should try testing 2.6.18 kernel, to the point of adding an extra disk or reinstalling your OS. -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Pasi Kärkkäinen
2010-Jan-18 06:42 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 10:47:47AM +0700, Fajar A. Nugraha wrote:> On Mon, Jan 18, 2010 at 3:23 AM, mail ignored <0.bugs.only.0@gmail.com> wrote: > > Is continuing to work with/on getting RHEL/CentOS 5 kernel-xen to boot > > of any particular use/value to _this_ issue, given the successful > > build/boot on xen.git''s pvops kernel (results posted above)? > > It''s basically like this: The "stable", officially supported kernel > version for xen as of now is still 2.6.18. When you encounter possible > kernel-related problems, it''s usually best to check if the problem is > present when using that kernel version. If it works, then you have one > quick workaorund: simply use 2.6.18 kernel. I recommended RHEL/Centos5 > kernel since it''s well maintained in terms of security fixes. >While http://xenbits.xen.org/linux-2.6.18-xen.hg (the official dom0 kernel for up to Xen 3.4.x) might have more xen-related fixes compared to RHEL/CentOS5 kernel-xen :) -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha
2010-Jan-18 07:22 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 1:42 PM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> On Mon, Jan 18, 2010 at 10:47:47AM +0700, Fajar A. Nugraha wrote: >> On Mon, Jan 18, 2010 at 3:23 AM, mail ignored <0.bugs.only.0@gmail.com> wrote: >> > Is continuing to work with/on getting RHEL/CentOS 5 kernel-xen to boot >> > of any particular use/value to _this_ issue, given the successful >> > build/boot on xen.git''s pvops kernel (results posted above)? >> >> It''s basically like this: The "stable", officially supported kernel >> version for xen as of now is still 2.6.18. When you encounter possible >> kernel-related problems, it''s usually best to check if the problem is >> present when using that kernel version. If it works, then you have one >> quick workaorund: simply use 2.6.18 kernel. I recommended RHEL/Centos5 >> kernel since it''s well maintained in terms of security fixes. >> > > While http://xenbits.xen.org/linux-2.6.18-xen.hg (the official dom0 > kernel for up to Xen 3.4.x) might have more xen-related fixes compared > to RHEL/CentOS5 kernel-xen :)correct :D However, after spending over three hours, opensuse 11.2 and kernel 2.6.18 is simply a no-no combo :P opensuse''s udev requires newer kernel, and using kernel 2.6.18 gives "udevd: error getting signalfd". Since udevd refuses to start, md wont start, lvm won''t start, and so on, and so on :( -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Keir Fraser
2010-Jan-18 07:27 UTC
Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On 17/01/2010 22:21, "mail ignored" <0.bugs.only.0@gmail.com> wrote:>> Possibly Xen''s >> ''clocksource='' boot option could be used to select a different platform >> timer exhibiting less jitter. > > currently, i have " ... clocksource=xen ... "I meant clocksource= on Xen''s command line, not dom0''s. I don''t think it even has an effect when specified to a kernel running on Xen. For example, you could try clocksource=acpi, or clocksource=pit. And then check that the ''Platform timer is...'' line from Xen boot output changes correspondingly. -- Keir _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-18 17:21 UTC
[Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi On Sun, Jan 17, 2010 at 11:27 PM, Keir Fraser <keir.fraser@eu.citrix.com> wrote:> I meant clocksource= on Xen''s command line, not dom0''s. I don''t think it > even has an effect when specified to a kernel running on Xen.where is that documented? tbh, i''ve never noticed those options at all. checking here, http://wiki.xensource.com/xenwiki/?action=fullsearch&context=180&value=clocksource&fullsearch=Text those references only mention clocksource= @kernel line, and the xen user''s manual doesn''t mention the term clocksource at all :-/ there''s no mention> For example, you could try clocksource=acpi, or clocksource=pit. And then > check that the ''Platform timer is...'' line from Xen boot output changes > correspondingly.checking those now ... will report results in a bot. thanks. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
mail ignored
2010-Jan-18 18:19 UTC
[Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 9:21 AM, mail ignored <0.bugs.only.0@gmail.com> wrote:>> For example, you could try clocksource=acpi, or clocksource=pit. And then >> check that the ''Platform timer is...'' line from Xen boot output changes >> correspondingly. > > checking those now ... will report results in a bit.both seem to work ... booting to, /xen.gz ... clocksource=acpi ... then, xm dmesg | grep -i timer (XEN) ACPI: PM-Timer IO Port: 0x808 (XEN) ..TIMER: vector=0xF0 apic1=0 pin1=2 apic2=-1 pin2=-1 (XEN) Platform timer is 3.579MHz ACPI PM Timer (XEN) HPET: 4 timers in total, 3 timers will be used for broadcast (XEN) mcheck_poll: Machine check polling timer started. ~ 15 minutes, quiet/stable operation @ stratum=2, full reach. assID=0 status=0684 leap_none, sync_ntp, 8 events, event_peer/strat_chg, version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.8-0.1-xen", leap=00, stratum=2, precision=-20, rootdelay=13.970, rootdispersion=26.079, peer=33500, refid=216.218.254.202, reftime=ceff1d2c.ddbe13c0 Mon, Jan 18 2010 9:47:24.866, poll=6, clock=ceff1e07.464f2053 Mon, Jan 18 2010 9:51:03.274, state=4, offset=15.245, frequency=24.962, jitter=10.571, noise=17.706, stability=7.306 remote refid st t when poll reach delay offset jitter ============================================================================= -clock.fmt.he.ne .PPS. 1 u 41 64 377 16.047 20.346 10.336 -otc2.psu.edu 130.207.244.240 2 u 39 64 377 100.626 14.538 5.387 +clock.isc.org .GPS. 1 u 54 64 377 15.174 -0.843 10.578 *clock.sjc.he.ne .CDMA. 1 u 26 64 377 13.970 12.255 7.051 zorro.sf-bay.or 216.218.254.202 2 u 31 64 377 16.489 13.730 5.781 +nist1.aol-ca.tr .ACTS. 1 u 15 64 377 14.891 5.947 14.596 booting to, /xen.gz ... clocksource=pit ... xm dmesg | grep -i timer (XEN) ACPI: PM-Timer IO Port: 0x808 (XEN) ..TIMER: vector=0xF0 apic1=0 pin1=2 apic2=-1 pin2=-1 (XEN) Platform timer is 1.193MHz PIT (XEN) mcheck_poll: Machine check polling timer started. ~ 20 minutes, quiet/stable operation @ stratum=2, full reach. assID=0 status=0654 leap_none, sync_ntp, 5 events, event_peer/strat_chg, version="ntpd 4.2.4p8@1.1612 Mon Dec 28 17:26:30 UTC 2009 (1)", processor="x86_64", system="Linux/2.6.31.8-0.1-xen", leap=00, stratum=2, precision=-20, rootdelay=16.981, rootdispersion=31.958, peer=63536, refid=207.200.81.113, reftime=ceff2377.503b4ceb Mon, Jan 18 2010 10:14:15.313, poll=6, clock=ceff240e.fa89c747 Mon, Jan 18 2010 10:16:46.978, state=4, offset=9.641, frequency=4.070, jitter=12.958, noise=6.244, stability=0.834 remote refid st t when poll reach delay offset jitter ============================================================================= +clock.fmt.he.ne .PPS. 1 u 30 64 177 15.250 8.422 9.657 +otc2.psu.edu 147.84.59.145 2 u 33 64 177 99.375 10.359 8.906 +clock.isc.org .GPS. 1 u 25 64 177 13.060 2.679 9.501 +clock.sjc.he.ne .CDMA. 1 u 22 64 177 15.097 10.608 7.866 +zorro.sf-bay.or 216.218.254.202 2 u 27 64 177 15.082 9.639 8.443 *nist1.aol-ca.tr .ACTS. 1 u 23 64 177 16.981 15.780 7.743 seemingly, _nothing_ to do with the kernel, then. rather overriding whatever xen''s default clocksource is does the trick. i''ll still take a look over longer term @ Dom0, and see if there''s any strangeness in DomU time ... thanks! _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
mail ignored
2010-Jan-18 18:26 UTC
[Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
wrt acpi vs pit, reading @ http://the-b.org/Linux_timers, it seems (?) acpi may be, in general, the better options, as pit " ... doesn''t gain lost time because it doesn''t have lost tick correction code ..." iiuc, that references those timers @kernel -- not clear if it is/isn''t relevant to Xen ... _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Pasi Kärkkäinen
2010-Jan-18 19:29 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 10:26:10AM -0800, mail ignored wrote:> wrt acpi vs pit, > > reading @ http://the-b.org/Linux_timers, it seems (?) acpi may be, in > general, the better options, as pit " ... doesn''t gain lost time > because it doesn''t have lost tick correction code ..." > > iiuc, that references those timers @kernel -- not clear if it is/isn''t > relevant to Xen ... >What was the default clocksource for you, showing the bad ntpd behaviour? -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
mail ignored
2010-Jan-18 21:50 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 11:29 AM, Pasi Kärkkäinen <pasik@iki.fi> wrote:> What was the default clocksource for you, showing the bad ntpd > behaviour?from above, > On Sun, Jan 17, 2010 at 1:33 PM, Keir Fraser <keir.fraser@eu.citrix.com> wrote: >> What platform timer does Xen say is initialised during boot? > > xm dmesg | egrep -i "timer|clock" > (XEN) ACPI: PM-Timer IO Port: 0x808 > (XEN) ..TIMER: vector=0xF0 apic1=0 pin1=2 apic2=-1 pin2=-1 > (XEN) Platform timer is 14.284MHz HPET > (XEN) HPET: 4 timers in total, 3 timers will be used for broadcast > (XEN) mcheck_poll: Machine check polling timer started. , looks like HPET, _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
mail ignored
2010-Jan-19 02:12 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
this page, http://wiki.xensource.com/xenwiki/xenpm seems to address hpet, pit & acpi timer sources. and does so without every mentioning ''clocksource'' ... tbh, i can''t get past the broken english on the page; it would be really helpful if someone who actually understands that stuff cleaned up the readability. the recommendations there are( i think?) "by far" to use HPET/PIT. which contraindicates the conclusion that I came to above :-/ so, questions are: (1) *is* there, in fact, a bug here? (2) which Timer is, in keeping with any/all issues discussed @ the aforementioned wiki page, recommended? (3) is the *real* answer simply intended to be "try them all, pick what works"? _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Pasi Kärkkäinen
2010-Jan-19 07:17 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Mon, Jan 18, 2010 at 06:12:44PM -0800, mail ignored wrote:> this page, > > http://wiki.xensource.com/xenwiki/xenpm > > seems to address hpet, pit & acpi timer sources. and does so without > every mentioning ''clocksource'' ... > tbh, i can''t get past the broken english on the page; it would be > really helpful if someone who actually understands that stuff cleaned > up the readability. >I added a couple of "clocksource" words and added one chapter to the end of the page. -- Pasi _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2010-Jan-19 09:24 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On 19/01/2010 02:12, "mail ignored" <0.bugs.only.0@gmail.com> wrote:> (1) *is* there, in fact, a bug here?It''s not clear. It could alternatively be a problem with the HPET on that specific system. Have you reproduced your issues on a different type of system (making sure that HPET platform timer is detected and used on that system too, of course)? That would then obviously point more at it being a software bug. What is the make/model of the system exhibiting the problem, by the way?> (2) which Timer is, in keeping with any/all issues discussed @ the > aforementioned wiki page, recommended?HPET or ACPI-PM timers ought to work equally well, assuming bugs haven''t crept into Xen''s code paths for either one, and both are implemented correctly by the hardware Xen is running on. I have to say I think it''s unlikely to be a software bug, because Xen''s code paths to use HPET and ACPI-PM timers are really very simple indeed -- it''s hard to see where a bug could exist in there. But of course that doesn''t make it absolutely impossible.> (3) is the *real* answer simply intended to be "try them all, pick what > works"?Well, really there oughtn''t to be enough of a difference between them for anyone to really notice or care which is actually being used. If this does turn out to be a hardware issue, then yes that is just the kind of case where you might expect to have to tweak things on the Xen command line. One option for us is to change the priority ordering of platform timers to prefer ACPI-PM above HPET, when both are available. That would assume that ACPI-PM is more reliable than HPET across a wide range of systems though. We don''t really have much evidence one way or the other. Another would be to dynamically detect platform timers that aren''t working well and disable them at run time. That would require some more thought, and isn''t going to be something for Xen 4.0 at this point. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
mail ignored
2010-Jan-19 17:58 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
hi On Tue, Jan 19, 2010 at 1:24 AM, Keir Fraser <keir.fraser@eu.citrix.com> wrote:> On 19/01/2010 02:12, "mail ignored" <0.bugs.only.0@gmail.com> wrote: > >> (1) *is* there, in fact, a bug here? > > It''s not clear. It could alternatively be a problem with the HPET on that > specific system. Have you reproduced your issues on a different type of > system (making sure that HPET platform timer is detected and used on that > system too, of course)? That would then obviously point more at it being a > software bug. What is the make/model of the system exhibiting the problem, > by the way?the 5 boxes on which this is currently tested/reproducible are mobo + cpu: (1)(2) ASUS M4A78T-E AM3 AMD 790GX ATX, 2 @ AMD Phenom II X4 945 cpu (3) MSI 790FX-GD70 AM3 AMD 790FX ATX, 1 @ AMD Phenom II X4 920 cpu (4)(5) GIGABYTE GA-MA770T-UD3P AM3 AMD 770 ATX, 1 @ AMD Phenom II X4 945 cpu 1 @ AMD Phenom II X4 920 cpu all opensuse 11.2 ...>> (2) which Timer is, in keeping with any/all issues discussed @ the >> aforementioned wiki page, recommended? > > HPET or ACPI-PM timers ought to work equally well, assuming bugs haven''t > crept into Xen''s code paths for either one, and both are implemented > correctly by the hardware Xen is running on. I have to say I think it''s > unlikely to be a software bug, because Xen''s code paths to use HPET and > ACPI-PM timers are really very simple indeed -- it''s hard to see where a bug > could exist in there. But of course that doesn''t make it absolutely > impossible.from reading the wiki page, i got the sense that HPET is "by far" recommended -- if only due to lack of specific mention of acpi(_pm) as clocksource. can''t argue with the fact that it works, though. if "ACPI-PM timers ought to work equally well", then I''ve a clear option/solution -- and knowledge enuf, now, to know where to look for related problems should they creep up in the future, perhaps on new/different hardware.>> (3) is the *real* answer simply intended to be "try them all, pick what >> works"? > > Well, really there oughtn''t to be enough of a difference between them for > anyone to really notice or care which is actually being used.ok. though, reproducibly across all the aforementioned systems, there is.> If this does > turn out to be a hardware issue, then yes that is just the kind of case > where you might expect to have to tweak things on the Xen command line.the ''common'' hardware is, of course, the AMD Phenom II X4 cpu ...> One option for us is to change the priority ordering of platform timers to > prefer ACPI-PM above HPET, when both are available. That would assume that > ACPI-PM is more reliable than HPET across a wide range of systems though. We > don''t really have much evidence one way or the other.that really only makes sense, doesn''t it, if it''s known to be NOT problematic on other platforms. given my obviously limited expertise in the xen source, this all ''smells'' like a problem ...> Another would be to dynamically detect platform timers that aren''t working > well and disable them at run time. That would require some more thought, and > isn''t going to be something for Xen 4.0 at this point.understood. pls let me know if there''s additional info you need; or, if this should issue should simply be allowed to ''fade'' now that there''s a clear WORKSFORSOME solution. thanks. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2010-Jan-19 18:04 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On 19/01/2010 17:58, "mail ignored" <0.bugs.only.0@gmail.com> wrote:>> If this does >> turn out to be a hardware issue, then yes that is just the kind of case >> where you might expect to have to tweak things on the Xen command line. > > the ''common'' hardware is, of course, the AMD Phenom II X4 cpu ...The particular hardware that really matters is the chipset, as that is where the HPET resides. Is that common across all the systems? -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
mail ignored
2010-Jan-19 18:24 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On Tue, Jan 19, 2010 at 10:04 AM, Keir Fraser <keir.fraser@eu.citrix.com> wrote:> On 19/01/2010 17:58, "mail ignored" <0.bugs.only.0@gmail.com> wrote: > >>> If this does >>> turn out to be a hardware issue, then yes that is just the kind of case >>> where you might expect to have to tweak things on the Xen command line. >> >> the ''common'' hardware is, of course, the AMD Phenom II X4 cpu ... > > The particular hardware that really matters is the chipset, as that is where > the HPET resides. Is that common across all the systems? >no, not identical, (1)(2) ASUS M4A78T-E North Bridge AMD 790GX South Bridge AMD SB750 (3) MSI 790FX-GD70 North Bridge AMD 790FX South Bridge AMD SB750 (4)(5) GIGABYTE GA-MA770T-UD3P North Bridge AMD 770 South Bridge AMD SB710 but, perhaps, related enough? _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2010-Jan-19 18:35 UTC
Re: [Xen-devel] Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.
On 19/01/2010 18:04, "Keir Fraser" <keir.fraser@eu.citrix.com> wrote:>>> If this does >>> turn out to be a hardware issue, then yes that is just the kind of case >>> where you might expect to have to tweak things on the Xen command line. >> >> the ''common'' hardware is, of course, the AMD Phenom II X4 cpu ... > > The particular hardware that really matters is the chipset, as that is where > the HPET resides. Is that common across all the systems?Oh, I see they do vary, although all are AMD chipsets of presumably a similar vintage. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel