Re: [Xen-users] ntpd under Xen Dom0 exhibits extremely high jitter/noise? runs stable/quiet under non-xen kernel.

[mail ignored <0.bugs.only.0@xxxxxxxxx>]

On Wed, Jan 13, 2010 at 2:10 PM, mail ignored <0.bugs.only.0@xxxxxxxxx> 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@xxxxxxxx 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@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users