All -
This is going to be a long email. My apologies for that! I have a
question regarding performance of OpenNTPD vs generic/DaveMills NTPD.
Setup:
1. We have 3 machines in our DMZ which act as our primary NTP servers.
2. We have close to 8 machines in our Intranet which act as secondary
NTP servers, which serve all our network and they all get their
time(rather try to) from our Primary NTP servers.
All the 3 primary NTP servers run OpenBSD 3.9 stable (i.e from the CD)
and GENERIC kernel. and are in DMZ. All the them have identical
hardware, in fact Dell PowerEdge 650 and all have same setup,
* same release install i.e 3.9 install from OpenBSD CD.
* GENERIC kernel and NO modification using ukc or recompile
DMZ NTP servers.
dmz-ntp1
uname: OpenBSD dmz-ntp1.XXX.YYY 3.9 GENERIC#617 i386
Snippet of dmesg:
OpenBSD 3.9 (GENERIC) #617: Thu Mar 2 02:26:48 MST 2006
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
dmz-ntp2
uname: OpenBSD dmz-ntp2.XXX.YYY 3.9 GENERIC#617 i386
OpenBSD 3.9 (GENERIC) #617: Thu Mar 2 02:26:48 MST 2006
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
dmz-ntp3
uname: OpenBSD dmz-ntp3 XXX.YYY 3.9 GENERIC#617 i386
OpenBSD 3.9 (GENERIC) #617: Thu Mar 2 02:26:48 MST 2006
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
ntpd configuration is the same for all the DMZ boxes.
Shown below are with comments removed!
[Begin Conf]
server 0.us.pool.ntp.org
server 1.us.pool.ntp.org
server 2.us.pool.ntp.org
server dmz-ntp1.XXX.YYY
server dmz-ntp2.XXX.YYY
server dmz-ntp3.XXX.YYY
listen on *
[End Conf]
All our internal boxes run RHEL4u2
Linux arrowhead.XXX.YYY 2.6.9-22.ELsmp #1 SMP Mon Sep 19 18:32:14 EDT
2005 i686 i686 i386 GNU/Linux
ntpd configuration is the same for all internal boxes.
Show below are with comments removed.
[Begin Conf]
server dmz-ntp1.XXX.YYY
server dmz-ntp2.XXX.YYY
server dmz-ntp3.XXX.YYY
peer ntp01.XXX.YYY
peer ntp02.XXX.YYY
peer ntp03.XXX.YYY
peer ntp04.XXX.YYY
peer ntp05.XXX.YYY
peer ntp06.XXX.YYY
peer ntp07.XXX.YYY
peer ntp08.XXX.YYY
peer ntp09.XXX.YYY
peer ntp10.XXX.YYY
server 127.127.1.0
fudge 127.127.1.0 stratum 10
driftfile /var/lib/ntp/drift
[End Conf]
All NTP machines are reachable on port 123.
Problem:
There seems to be a lot of jitter on OpenNTPD based machines (i.e DMZ
machines). This prevents the Intranet machines to sync up to the OpenBSD
machines. So we installed NTPD (not OpenNTPD) from ports and restared
NTP on the dmz-ntp3 and lo and behold all Linux boxes starting syncing
up with this box, but not to any other OpenNTPD based machines (which
are 2 right now).
NTP stats from the Intranet boxes.
Please ignore
NTP stats from our Intranet box called arrowhead running RHEL4u2
--== ntpq -p arrowhead ==--
remote refid st t when poll reach delay offset jitter
==============================================================================
dmz-ntp1 93.5.230.181 4 u 22 1024 377 0.734 -270.18 87.034
-dmz-ntp3 24.123.214.97 3 u 903 1024 377 1.373 -10.758 3.727
xdmz-ntp2 220.249.119.159 4 u 960 1024 377 0.932 -213.16 246.533
arrowhead .STEP. 16 u - 1024 0 0.000 0.000 4000.00
+aspen 10.200.2.65 5 u 559 1024 376 4.154 -0.673 0.142
+baldy 10.200.1.243 5 u 270 1024 377 3.364 0.145 1.597
buttermilk 10.200.2.87 5 u 576 1024 376 2.387 0.718 0.203
-copper 10.200.1.222 4 u 971 1024 376 3.130 1.433 0.584
cypress 10.200.2.87 5 u 426 1024 376 4.032 -0.456 0.111
*heavenly 10.200.1.222 4 u 772 1024 376 3.314 -0.136 0.630
-kirkwood 10.200.2.68 3 u 610 1024 377 4.017 -2.248 3.995
-shasta 66.92.68.11 2 u 747 1024 376 10.674 -12.836 3.089
LOCAL(0) LOCAL(0) 10 l 63 64
NTP stats from our Intranet box called aspen running RHEL4u2
--== ntpq -p aspen ==--
remote refid st t when poll reach delay offset jitter
==============================================================================
xdmz-ntp1 93.5.230.181 4 u 292 1024 377 0.867 -240.39 86.761
-dmz-ntp3 24.123.214.97 3 u 797 1024 377 1.143 -12.227 3.712
xdmz-ntp2 220.249.119.159 4 u 908 1024 377 2.641 -222.98 247.827
-arrowhead 10.200.1.245 5 u 552 1024 377 3.845 0.828 0.409
aspen .STEP. 16 u - 1024 0 0.000 0.000 4000.00
-baldy 10.200.1.222 4 u 1015 1024 376 4.710 -0.608 0.941
+buttermilk 10.200.1.245 5 u 888 1024 376 3.913 0.013 0.210
*copper 10.200.1.222 4 u 979 1024 376 6.086 0.130 0.248
-cypress 10.200.2.87 5 u 989 1024 376 3.789 0.254 0.437
-heavenly 10.200.2.87 5 u 299 1024 377 5.326 1.349 0.759
+kirkwood 199.184.165.135 3 u 284 1024 377 9.448 -0.471 0.523
-shasta 66.33.216.11 3 u 26 1024 376 5.536 -16.241 4.331
LOCAL(0) LOCAL(0) 10 l 52 64 377 0.000 0.000 0.001
NTP stats from our Intranet box called baldy running RHEL4u2
--== ntpq -p baldy ==--
remote refid st t when poll reach delay offset jitter
==============================================================================
LOCAL(0) LOCAL(0) 10 l 73 64 377 0.000 0.000
0.004
xdmz-ntp1 93.5.230.181 4 u 20 1024 377 0.712 -269.32 86.366
-dmz-ntp3 24.123.214.97 3 u 427 1024 377 0.741 -11.506 4.435
xdmz-ntp2 220.249.119.159 4 u 918 1024 377 0.704 -219.70 247.097
-arrowhead 10.200.1.222 4 u 620 1024 376 3.329 -0.162 1.737
+aspen 10.200.2.65 5 u 335 1024 377 3.949 0.988 0.622
baldy 0.0.0.0 16 u - 1024 0 0.000 0.000 4000.00
+buttermilk 10.200.1.231 5 u 80 1024 377 1.331 0.930 0.041
-copper 10.200.1.222 4 u 969 1024 376 3.394 2.795 0.740
-cypress 10.200.2.87 5 u 572 1024 376 2.765 0.420 2.524
*heavenly 10.200.1.222 4 u 684 1024 376 1.964 0.571 0.174
-kirkwood 199.184.165.135 3 u 321 1024 376 2.006 -1.500 1.571
-shasta 66.33.216.11 3 u 132 1024 377 2.625 -13.549 3.691
As you can see from the jitter column, the jitter for the dmz-ntp
servers are high as compared to other one (except for dmz-ntp3, which is
currently running DaveMills NTPD)
Question:
Why is OpenNTPD causing more jitter? Is there anybody experiencing this?
Is there a workaround? or does anybody need any more logs apart from
what I have provided?
To be needed?
I still have Intranet logs for all our Intranet ntp servers. I dont
think dmesg will be useful here, but anyways I will append a dmesg from
one of the dmz-ntp machines and dmesg is identical on all three dmz-ntp
machines.
dmesg from dmz-ntp1:
OpenBSD 3.9 (GENERIC) #617: Thu Mar 2 02:26:48 MST 2006
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel(R) Pentium(R) 4 CPU 3.06GHz ("GenuineIntel" 686-class) 3.07 GHz
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,SBF,CNXT-ID
real mem = 1073262592 (1048108K)
avail mem = 972619776 (949824K)
using 4278 buffers containing 53764096 bytes (52504K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(00) BIOS, date 06/02/03, BIOS32 rev. 0 @ 0xffe90
pcibios0 at bios0: rev 2.1 @ 0xf0000/0x10000
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xfc500/96 (4 entries)
pcibios0: no compatible PCI ICU found: ICU vendor 0x1166 product 0x0203
pcibios0: Warning, unable to fix up PCI interrupt routing
pcibios0: PCI bus #0 is the last bus
bios0: ROM list: 0xc0000/0x8000 0xc8000/0x4800 0xcc800/0x1000
0xec000/0x4000!
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "ServerWorks GCNB-LE Host" rev 0x32
pchb1 at pci0 dev 0 function 1 "ServerWorks GCNB-LE Host" rev 0x00
pci1 at pchb1 bus 1
em0 at pci1 dev 3 function 0 "Intel PRO/1000MT (82546EB)" rev 0x01: irq
10, address 00:04:23:a6:47:16
em1 at pci1 dev 3 function 1 "Intel PRO/1000MT (82546EB)" rev 0x01: irq
7, address 00:04:23:a6:47:17
vga1 at pci0 dev 4 function 0 "ATI Rage XL" rev 0x27
wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
pciide0 at pci0 dev 5 function 0 "CMD Technology PCI0680" rev 0x02
pciide0: bus-master DMA support present
pciide0: channel 0 wired to native-PCI mode
pciide0: using irq 15 for native-PCI interrupt
wd0 at pciide0 channel 0 drive 0: <ST3120026A>
wd0: 16-sector PIO, LBA48, 114473MB, 234441648 sectors
wd0(pciide0:0:0): using PIO mode 4, Ultra-DMA mode 5
pciide0: channel 1 wired to native-PCI mode
piixpm0 at pci0 dev 15 function 0 "ServerWorks CSB6" rev 0xa0: SMBus
disabled
pciide1 at pci0 dev 15 function 1 "ServerWorks CSB6 RAID/IDE" rev 0xa0: DMA
atapiscsi0 at pciide1 channel 0 drive 0
scsibus0 at atapiscsi0: 2 targets
cd0 at scsibus0 targ 0 lun 0: <TEAC, CD-224E, K.9A> SCSI0 5/cdrom removable
cd0(pciide1:0:0): using PIO mode 4, DMA mode 2, Ultra-DMA mode 2
pciide1: no compatibility interrupt for use by channel 1
ohci0 at pci0 dev 15 function 2 "ServerWorks CSB6 USB" rev 0x05: irq 11,
version 1.0, legacy support
usb0 at ohci0: USB revision 1.0
uhub0 at usb0
uhub0: ServerWorks OHCI root hub, rev 1.00/1.00, addr 1
uhub0: 4 ports with 4 removable, self powered
pcib0 at pci0 dev 15 function 3 "ServerWorks GCLE-2 Host" rev 0x00
pchb2 at pci0 dev 16 function 0 "ServerWorks CIOB-E" rev 0x12
pchb3 at pci0 dev 16 function 2 "ServerWorks CIOB-E" rev 0x12
pci2 at pchb3 bus 2
isa0 at pcib0
isadma0 at isa0
pckbc0 at isa0 port 0x60/5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd slot
wskbd0 at pckbd0: console keyboard, using wsdisplay0
pmsi0 at pckbc0 (aux slot)
pckbc0: using irq 12 for aux slot
wsmouse0 at pmsi0 mux 0
pcppi0 at isa0 port 0x61
midi0 at pcppi0: <PC speaker>
spkr0 at pcppi0
npx0 at isa0 port 0xf0/16: using exception 16
pccom0 at isa0 port 0x3f8/8 irq 4: ns16550a, 16 byte fifo
fdc0 at isa0 port 0x3f0/6 irq 6 drq 2
fd0 at fdc0 drive 0: 1.44MB 80 cyl, 2 head, 18 sec
biomask eb6d netmask efed ttymask ffef
pctr: user-level cycle counter enabled
dkcsum: wd0 matches BIOS drive 0x80
root on wd0a
rootdev=0x0 rrootdev=0x300 rawdev=0x302
