Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 22 Mar 2013 17:34:43 -0700 (PDT)
From:      Paul Pathiakis <pathiaki2@yahoo.com>
To:        Davide D'Amico <davide.damico@contactlab.com>, =?utf-8?B?0JXQstCz0LXQvdC40Lkg0KXQvtGA0LrQuNC9?= <khorkin.eugeniy@gmail.com>
Cc:        "freebsd-performance@freebsd.org" <freebsd-performance@freebsd.org>
Subject:   Re: FreeBSD 9.1 vs CentOS 6.3
Message-ID:  <1363998883.22604.YahooMailNeo@web141401.mail.bf1.yahoo.com>
In-Reply-To: <ebfc033f-d8db-4162-86be-db013e3d7eb2@email.android.com>
References:  <CAF3xD3nDDUYB94TS9AUUQ=CPztB1S8mU4fRwNB5GupmK8MgwXg@mail.gmail.com> <ebfc033f-d8db-4162-86be-db013e3d7eb2@email.android.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi,=0A=0AThere are several things about this that are highly suspect.=C2=A0=
 =0A=0A=0AFirst, wipe out the hardware RAID.=C2=A0 The processor doing RAID=
 computation is, probably, MUCH slower than a core on the CPU.=C2=A0 Even i=
f it's RAID-1 (Simple Mirror) this RAID card is performing tasks that is do=
es not need to do including replicating writes to two targets from the cont=
roller or checking it's cache, battery, etc.=C2=A0 If it's possible to disa=
ble the onboard cache, do it.=0A=0A=0A(This controller seems to use the mfi=
 driver so that's probably a re-badged LSI with an LSI200x chipset.=C2=A0 (=
Lately, I have not been too thrilled with the mfi driver performance.=C2=A0=
 The mpt seems much more stable and robust - my opinion not bashing to whoe=
ver wrote it.)=0A=0A=0ATake the card to full JBOD mode before testing your =
"throughput".=0A=0AAlso, make sure that all on-disk caches are "off".=C2=A0=
 You do not want to get into a double-buffer flush corruption issue.=0A=0AI=
 see you are using two SSDs as ARC2 caches....=C2=A0 Quite frankly, junk th=
em as well, for now.=C2=A0 You may want to deal with increasing throughput =
once you find out what the issue is with the base throughput.=0A=0A=0ACreat=
e the appropriate pools/mirrors as necessary, all on simple JBOD.=0A=0AFrom=
 there, run a simple test to see what your throughput is on single spindle.=
=C2=A0 (I'm talking SIMPLE.=C2=A0 dd if=3D/dev/zero of=3D/<path>/zeroes.out=
=C2=A0 bs=3D128k count=3D<some ridiculous number like 100000 or more> - Thi=
s should be over 100 MB/sec throughput.=C2=A0 (with those drives it may be =
150 MB/sec or more)=0A=0AAfter that, check the throughput of the card/chann=
el.=C2=A0 Write a very simple script script that performs the same throughp=
ut test on half the available spindles (something as ludicrous as the same =
dd command to half your spindles throwing it in the background as each kick=
s off - Then, do it for all of your drives)=C2=A0 =0A=0A=0AThis is an imper=
fect test but it's down and dirty.=C2=A0 You can use the performance/benchm=
ark tools later.=0A=0ALet's just remember that the throughput of 6 Gb/sec i=
s gigaBITS not gigaBYTES.=C2=A0 There's a factor of 8 difference.... or 750=
 MB/sec... theoretical.=0A=0AYou should exceed channel speed with just 5 of=
 these drives/spindles... use 6 to be sure (I would think you won't see agg=
regate throughput beyond 725 MB/sec....=C2=A0 If any tool is showing you th=
at on any OS with this configuration, they aren't doing proper calculations=
 and/or they are doing MB vs MiB.=C2=A0 =0A=0A=0AIf you can't get up to tha=
t speed, you may have a driver issue or channel issue.=0A=0A(I was getting =
as high as 700 MB/sec with 6 SATA drives doing it this way.... No SSD, ASUS=
 MB 990FX with 1100T 6 core proc, 16 GB RAM... )=0A=0AP.=0A=0A=0A=0A_______=
_________________________=0A From: Davide D'Amico <davide.damico@contactlab=
.com>=0ATo: =D0=95=D0=B2=D0=B3=D0=B5=D0=BD=D0=B8=D0=B9 =D0=A5=D0=BE=D1=80=
=D0=BA=D0=B8=D0=BD <khorkin.eugeniy@gmail.com> =0ACc: freebsd-performance@f=
reebsd.org =0ASent: Friday, March 22, 2013 12:15 PM=0ASubject: Re: FreeBSD =
9.1 vs CentOS 6.3=0A =0AWell, the I/O isn't the bottleneck (if you follow t=
he link to freebsd-fs, you'll see iostats values) but it seems something re=
lated to cpu/scheduler or something else.=0ANow I am trying vmware 5 on the=
 same server and a vm with centos6: the vm outperforms freebsd with every c=
oncurrency from 1 to 48 threads in sysbench oltp tests.=0A=0AThanks,=0Ad.=
=0A=0A"=D0=95=D0=B2=D0=B3=D0=B5=D0=BD=D0=B8=D0=B9 =D0=A5=D0=BE=D1=80=D0=BA=
=D0=B8=D0=BD" <khorkin.eugeniy@gmail.com> ha scritto:=0A=0A>Hi Davide!=0A>=
=0A>Are you sure that disk is the bottleneck in your test?=0A>Does systat -=
vm 1 show 100% busy for disk ?=0A>=0A>Evgeny.=0A>=0A>=0A>2013/3/22 Davide D=
'Amico <davide.damico@contactlab.com>=0A>=0A>> Hi, I'm doing performance te=
sts on a DELL R720, follows dmesg:=0A>>=0A>> Copyright (c) 1992-2012 The Fr=
eeBSD Project.=0A>> Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991,=
 1992, 1993,=0A>1994=0A>>=C2=A0 =C2=A0 =C2=A0 =C2=A0  The Regents of the Un=
iversity of California. All rights=0A>reserved.=0A>> FreeBSD is a registere=
d trademark of The FreeBSD Foundation.=0A>> FreeBSD 9.1-RELEASE #0: Tue Mar=
 19 10:24:21 CET 2013=0A>>=C2=A0 =C2=A0  root:/usr/obj/usr/src/sys/R720 amd=
64=0A>> CPU: Intel(R) Xeon(R) CPU E5-2640 0 @ 2.50GHz (2500.05-MHz K8-class=
=0A>CPU)=0A>>=C2=A0  Origin =3D "GenuineIntel"=C2=A0 Id =3D 0x206d7=C2=A0 F=
amily =3D 6=C2=A0 Model =3D 2d=0A>Stepping =3D=0A>> 7=0A>> Features=3D0xbfe=
bfbff<FPU,VME,**DE,PSE,TSC,MSR,PAE,MCE,CX8,**=0A>> APIC,SEP,MTRR,PGE,MCA,CM=
OV,**PAT,PSE36,CLFLUSH,DTS,ACPI,**=0A>> MMX,FXSR,SSE,SSE2,SS,HTT,TM,**PBE>=
=0A>> Features2=3D0x1fbee3ff<SSE3,**PCLMULQDQ,DTES64,MON,DS_CPL,**=0A>> VMX=
,SMX,EST,TM2,SSSE3,CX16,**xTPR,PDCM,PCID,DCA,SSE4.1,**=0A>> SSE4.2,x2APIC,P=
OPCNT,TSCDLT,**AESNI,XSAVE,OSXSAVE,AVX>=0A>>=C2=A0  AMD Features=3D0x2c1008=
00<SYSCALL,**NX,Page1GB,RDTSCP,LM>=0A>>=C2=A0  AMD Features2=3D0x1<LAHF>=0A=
>>=C2=A0  TSC: P-state invariant, performance statistics=0A>> real memory=
=C2=A0 =3D 34359738368 (32768 MB)=0A>> avail memory =3D 33027436544 (31497 =
MB)=0A>> Event timer "LAPIC" quality 600=0A>> ACPI APIC Table: <DELL=C2=A0 =
 PE_SC3=C2=A0 >=0A>> FreeBSD/SMP: Multiprocessor System Detected: 8 CPUs=0A=
>> FreeBSD/SMP: 2 package(s) x 2 core(s) x 2 SMT threads=0A>>=C2=A0 cpu0 (B=
SP): APIC ID:=C2=A0 0=0A>>=C2=A0 cpu1 (AP): APIC ID:=C2=A0 1=0A>>=C2=A0 cpu=
2 (AP): APIC ID: 10=0A>>=C2=A0 cpu3 (AP): APIC ID: 11=0A>>=C2=A0 cpu4 (AP):=
 APIC ID: 32=0A>>=C2=A0 cpu5 (AP): APIC ID: 33=0A>>=C2=A0 cpu6 (AP): APIC I=
D: 42=0A>>=C2=A0 cpu7 (AP): APIC ID: 43=0A>> ioapic1: Changing APIC ID to 1=
=0A>> ioapic2: Changing APIC ID to 2=0A>> ioapic0 <Version 2.0> irqs 0-23 o=
n motherboard=0A>> ioapic1 <Version 2.0> irqs 32-55 on motherboard=0A>> ioa=
pic2 <Version 2.0> irqs 64-87 on motherboard=0A>> kbd1 at kbdmux0=0A>> ctl:=
 CAM Target Layer loaded=0A>> acpi0: <DELL PE_SC3> on motherboard=0A>> acpi=
0: Power Button (fixed)=0A>> acpi0: reservation of 0, 2000 (3) failed=0A>> =
cpu0: <ACPI CPU> on acpi0=0A>> cpu1: <ACPI CPU> on acpi0=0A>> cpu2: <ACPI C=
PU> on acpi0=0A>> cpu3: <ACPI CPU> on acpi0=0A>> cpu4: <ACPI CPU> on acpi0=
=0A>> cpu5: <ACPI CPU> on acpi0=0A>> cpu6: <ACPI CPU> on acpi0=0A>> cpu7: <=
ACPI CPU> on acpi0=0A>> atrtc0: <AT realtime clock> port 0x70-0x7f irq 8 on=
 acpi0=0A>> Event timer "RTC" frequency 32768 Hz quality 0=0A>> attimer0: <=
AT timer> port 0x40-0x5f irq 0 on acpi0=0A>> Timecounter "i8254" frequency =
1193182 Hz quality 0=0A>> Event timer "i8254" frequency 1193182 Hz quality =
100=0A>> Timecounter "ACPI-fast" frequency 3579545 Hz quality 900=0A>> acpi=
_timer0: <24-bit timer at 3.579545MHz> port 0x808-0x80b on acpi0=0A>> hpet0=
: <High Precision Event Timer> iomem 0xfed00000-0xfed003ff on=0A>acpi0=0A>>=
 Timecounter "HPET" frequency 14318180 Hz quality 950=0A>> Event timer "HPE=
T" frequency 14318180 Hz quality 550=0A>> pcib0: <ACPI Host-PCI bridge> por=
t 0xcf8-0xcff on acpi0=0A>> pci0: <ACPI PCI bus> on pcib0=0A>> pcib1: <ACPI=
 PCI-PCI bridge> irq 53 at device 1.0 on pci0=0A>> pci1: <ACPI PCI bus> on =
pcib1=0A>> igb0: <Intel(R) PRO/1000 Network Connection version - 2.3.4> por=
t=0A>> 0xfc80-0xfc9f mem 0xdcb00000-0xdcbfffff,**0xdcff0000-0xdcff3fff irq=
=0A>38 at=0A>> device 0.0 on pci1=0A>> igb0: Using MSIX interrupts with 9 v=
ectors=0A>> igb0: Ethernet address: bc:30:5b:ef:27:b8=0A>> [...]=0A>> igb1:=
 <Intel(R) PRO/1000 Network Connection version - 2.3.4> port=0A>> 0xfca0-0x=
fcbf mem 0xdcc00000-0xdccfffff,**0xdcff4000-0xdcff7fff irq=0A>37 at=0A>> de=
vice 0.1 on pci1=0A>> igb1: Using MSIX interrupts with 9 vectors=0A>> igb1:=
 Ethernet address: bc:30:5b:ef:27:b9=0A>> [...]=0A>> igb2: <Intel(R) PRO/10=
00 Network Connection version - 2.3.4> port=0A>> 0xfcc0-0xfcdf mem 0xdcd000=
00-0xdcdfffff,**0xdcff8000-0xdcffbfff irq=0A>36 at=0A>> device 0.2 on pci1=
=0A>> igb2: Using MSIX interrupts with 9 vectors=0A>> igb2: Ethernet addres=
s: bc:30:5b:ef:27:ba=0A>> [...]=0A>> igb3: <Intel(R) PRO/1000 Network Conne=
ction version - 2.3.4> port=0A>> 0xfce0-0xfcff mem 0xdce00000-0xdcefffff,**=
0xdcffc000-0xdcffffff irq=0A>34 at=0A>> device 0.3 on pci1=0A>> igb3: Using=
 MSIX interrupts with 9 vectors=0A>> igb3: Ethernet address: bc:30:5b:ef:27=
:bb=0A>> [...]=0A>> pcib2: <ACPI PCI-PCI bridge> irq 53 at device 2.0 on pc=
i0=0A>> pci3: <ACPI PCI bus> on pcib2=0A>> pcib3: <ACPI PCI-PCI bridge> irq=
 53 at device 2.2 on pci0=0A>> pci2: <ACPI PCI bus> on pcib3=0A>> mfi0: <De=
ll PERC H710P Mini (monolithics)> port 0xec00-0xecff mem=0A>> 0xddffc000-0x=
ddffffff,**0xddf80000-0xddfbffff irq 42 at device 0.0 on=0A>pci2=0A>> mfi0:=
 Using MSI=0A>> mfi0: Megaraid SAS driver Ver 4.23=0A>> mfi0: MaxCmd =3D 3f=
0 MaxSgl =3D 46 state =3D b73c03f0=0A>> pcib4: <ACPI PCI-PCI bridge> irq 53=
 at device 3.0 on pci0=0A>> pci4: <ACPI PCI bus> on pcib4=0A>> pcib5: <ACPI=
 PCI-PCI bridge> irq 53 at device 3.2 on pci0=0A>> pci5: <ACPI PCI bus> on =
pcib5=0A>> pci0: <base peripheral> at device 5.0 (no driver attached)=0A>> =
pci0: <base peripheral> at device 5.2 (no driver attached)=0A>> pcib6: <PCI=
-PCI bridge> irq 16 at device 17.0 on pci0=0A>> pci6: <PCI bus> on pcib6=0A=
>> pci0: <simple comms> at device 22.0 (no driver attached)=0A>> pci0: <sim=
ple comms> at device 22.1 (no driver attached)=0A>> ehci0: <EHCI (generic) =
USB 2.0 controller> mem 0xdf8fe000-0xdf8fe3ff=0A>irq=0A>> 23 at device 26.0=
 on pci0=0A>> usbus0: EHCI version 1.0=0A>> usbus0 on ehci0=0A>> pcib7: <AC=
PI PCI-PCI bridge> at device 28.0 on pci0=0A>> device_attach: pcib7 attach =
returned 6=0A>> pcib7: <ACPI PCI-PCI bridge> irq 19 at device 28.7 on pci0=
=0A>> pci8: <ACPI PCI bus> on pcib7=0A>> pcib8: <PCI-PCI bridge> at device =
0.0 on pci8=0A>> pci9: <PCI bus> on pcib8=0A>> pcib9: <PCI-PCI bridge> at d=
evice 0.0 on pci9=0A>> pci10: <PCI bus> on pcib9=0A>> pcib10: <PCI-PCI brid=
ge> at device 0.0 on pci10=0A>> pci11: <PCI bus> on pcib10=0A>> vgapci0: <V=
GA-compatible display> mem 0xd8000000-0xd8ffffff,**=0A>> 0xdeffc000-0xdefff=
fff,**0xde000000-0xde7fffff irq 19 at device 0.0 on=0A>> pci11=0A>> pcib11:=
 <PCI-PCI bridge> at device 1.0 on pci9=0A>> pci12: <PCI bus> on pcib11=0A>=
> ehci1: <EHCI (generic) USB 2.0 controller> mem 0xdf8ff000-0xdf8ff3ff=0A>i=
rq=0A>> 22 at device 29.0 on pci0=0A>> usbus1: EHCI version 1.0=0A>> usbus1=
 on ehci1=0A>> pcib12: <PCI-PCI bridge> at device 30.0 on pci0=0A>> pci13: =
<PCI bus> on pcib12=0A>> isab0: <PCI-ISA bridge> at device 31.0 on pci0=0A>=
> isa0: <ISA bus> on isab0=0A>> pcib13: <ACPI Host-PCI bridge> on acpi0=0A>=
> pci64: <ACPI PCI bus> on pcib13=0A>> pcib14: <ACPI PCI-PCI bridge> irq 85=
 at device 1.0 on pci64=0A>> pci65: <ACPI PCI bus> on pcib14=0A>> pcib15: <=
ACPI PCI-PCI bridge> irq 85 at device 2.0 on pci64=0A>> pci66: <ACPI PCI bu=
s> on pcib15=0A>> pcib16: <ACPI PCI-PCI bridge> irq 85 at device 3.0 on pci=
64=0A>> pci67: <ACPI PCI bus> on pcib16=0A>> pcib17: <ACPI PCI-PCI bridge> =
irq 85 at device 3.2 on pci64=0A>> pci68: <ACPI PCI bus> on pcib17=0A>> pci=
64: <base peripheral> at device 5.0 (no driver attached)=0A>> pci64: <base =
peripheral> at device 5.2 (no driver attached)=0A>> pcib18: <ACPI Host-PCI =
bridge> on acpi0=0A>> pci63: <ACPI PCI bus> on pcib18=0A>> pci63: <base per=
ipheral> at device 8.0 (no driver attached)=0A>> [...]=0A>> pci63: <dasp> a=
t device 14.1 (no driver attached)=0A>> [...]=0A>> pci63: <dasp> at device =
19.1 (no driver attached)=0A>> pci63: <dasp> at device 19.4 (no driver atta=
ched)=0A>> pci63: <dasp> at device 19.5 (no driver attached)=0A>> pci63: <b=
ase peripheral> at device 19.6 (no driver attached)=0A>> pcib19: <ACPI Host=
-PCI bridge> on acpi0=0A>> pci127: <ACPI PCI bus> on pcib19=0A>> [...]=0A>>=
 pci127: <dasp> at device 14.1 (no driver attached)=0A>> [...]=0A>> pci127:=
 <dasp> at device 19.1 (no driver attached)=0A>> pci127: <dasp> at device 1=
9.4 (no driver attached)=0A>> pci127: <dasp> at device 19.5 (no driver atta=
ched)=0A>> pci127: <base peripheral> at device 19.6 (no driver attached)=0A=
>> uart1: <16550 or compatible> port 0x2f8-0x2ff irq 3 on acpi0=0A>> uart0:=
 <16550 or compatible> port 0x3f8-0x3ff irq 4 flags 0x10 on=0A>acpi0=0A>> o=
rm0: <ISA Option ROMs> at iomem 0xc0000-0xc7fff,0xec000-**0xeffff on=0A>isa=
0=0A>> sc0: <System console> at flags 0x100 on isa0=0A>> sc0: VGA <16 virtu=
al consoles, flags=3D0x300>=0A>> vga0: <Generic ISA VGA> at port 0x3c0-0x3d=
f iomem 0xa0000-0xbffff on=0A>isa0=0A>> est0: <Enhanced SpeedStep Frequency=
 Control> on cpu0=0A>> est: CPU supports Enhanced Speedstep, but is not rec=
ognized.=0A>> est: cpu_vendor GenuineIntel, msr 226d00001e00=0A>> device_at=
tach: est0 attach returned 6=0A>> p4tcc0: <CPU Frequency Thermal Control> o=
n cpu0=0A>> est1: <Enhanced SpeedStep Frequency Control> on cpu1=0A>> est: =
CPU supports Enhanced Speedstep, but is not recognized.=0A>> est: cpu_vendo=
r GenuineIntel, msr 226d00001e00=0A>> device_attach: est1 attach returned 6=
=0A>> p4tcc1: <CPU Frequency Thermal Control> on cpu1=0A>> est2: <Enhanced =
SpeedStep Frequency Control> on cpu2=0A>> est: CPU supports Enhanced Speeds=
tep, but is not recognized.=0A>> est: cpu_vendor GenuineIntel, msr 226d0000=
1e00=0A>> device_attach: est2 attach returned 6=0A>> p4tcc2: <CPU Frequency=
 Thermal Control> on cpu2=0A>> est3: <Enhanced SpeedStep Frequency Control>=
 on cpu3=0A>> est: CPU supports Enhanced Speedstep, but is not recognized.=
=0A>> est: cpu_vendor GenuineIntel, msr 224400001e00=0A>> device_attach: es=
t3 attach returned 6=0A>> p4tcc3: <CPU Frequency Thermal Control> on cpu3=
=0A>> est4: <Enhanced SpeedStep Frequency Control> on cpu4=0A>> est: CPU su=
pports Enhanced Speedstep, but is not recognized.=0A>> est: cpu_vendor Genu=
ineIntel, msr 226d00001e00=0A>> device_attach: est4 attach returned 6=0A>> =
p4tcc4: <CPU Frequency Thermal Control> on cpu4=0A>> est5: <Enhanced SpeedS=
tep Frequency Control> on cpu5=0A>> est: CPU supports Enhanced Speedstep, b=
ut is not recognized.=0A>> est: cpu_vendor GenuineIntel, msr 224400001e00=
=0A>> device_attach: est5 attach returned 6=0A>> p4tcc5: <CPU Frequency The=
rmal Control> on cpu5=0A>> est6: <Enhanced SpeedStep Frequency Control> on =
cpu6=0A>> est: CPU supports Enhanced Speedstep, but is not recognized.=0A>>=
 est: cpu_vendor GenuineIntel, msr 224400001e00=0A>> device_attach: est6 at=
tach returned 6=0A>> p4tcc6: <CPU Frequency Thermal Control> on cpu6=0A>> e=
st7: <Enhanced SpeedStep Frequency Control> on cpu7=0A>> est: CPU supports =
Enhanced Speedstep, but is not recognized.=0A>> est: cpu_vendor GenuineInte=
l, msr 224400001e00=0A>> device_attach: est7 attach returned 6=0A>> p4tcc7:=
 <CPU Frequency Thermal Control> on cpu7=0A>> mfi0: 6040 (417120269s/0x0020=
/info) - Shutdown command received from=0A>host=0A>> mfi0: 6041 (boot + 4s/=
0x0020/info) - Firmware initialization started=0A>(PCI=0A>> ID 005b/1000/1f=
34/1028)=0A>> mfi0: 6042 (boot + 4s/0x0020/info) - Firmware version 3.130.0=
5-1796=0A>> mfi0: 6043 (boot + 5s/0x0008/info) - Battery Present=0A>> mfi0:=
 6044 (boot + 5s/0x0020/info) - Package version 21.1.0-0007=0A>> mfi0: 6045=
 (boot + 5s/0x0020/info) - Board Revision A01=0A>> mfi0: 6046 (boot + 6s/0x=
0008/info) - Battery temperature is normal=0A>> mfi0: 6047 (boot + 6s/0x000=
8/info) - Current capacity of the battery=0A>is=0A>> above threshold=0A>> m=
fi0: 6048 (boot + 19s/0x0004/info) - Enclosure PD 20(c None/p1)=0A>> commun=
ication restored=0A>> mfi0: 6049 (boot + 19s/0x0002/info) - Inserted: Encl =
PD 20=0A>> mfi0: 6050 (boot + 19s/0x0002/info) - Inserted: PD 20(c None/p1)=
=0A>Info:=0A>> enclPd=3D20, scsiType=3Dd, portMap=3D00, sasAddr=3D5e4ae020a=
26cb500,**=0A>> 0000000000000000=0A>> mfi0: 6051 (boot + 19s/0x0002/info) -=
 Inserted: PD 00(e0x20/s0)=0A>> mfi0: 6052 (boot + 19s/0x0002/info) - Inser=
ted: PD 00(e0x20/s0) Info:=0A>> enclPd=3D20, scsiType=3D0, portMap=3D04, sa=
sAddr=3D5000c50054982339,**=0A>> 0000000000000000=0A>> mfi0: 6053 (boot + 1=
9s/0x0002/info) - Inserted: PD 01(e0x20/s1)=0A>> mfi0: 6054 (boot + 19s/0x0=
002/info) - Inserted: PD 01(e0x20/s1) Info:=0A>> enclPd=3D20, scsiType=3D0,=
 portMap=3D05, sasAddr=3D5000c50054986c91,**=0A>> 0000000000000000=0A>> mfi=
0: 6055 (boot + 19s/0x0002/info) - Inserted: PD 02(e0x20/s2)=0A>> mfi0: 605=
6 (boot + 19s/0x0002/info) - Inserted: PD 02(e0x20/s2) Info:=0A>> enclPd=3D=
20, scsiType=3D0, portMap=3D06, sasAddr=3D5000c50053da8f75,**=0A>> 00000000=
00000000=0A>> mfi0: 6057 (boot + 19s/0x0002/info) - Inserted: PD 03(e0x20/s=
3)=0A>> mfi0: 6058 (boot + 19s/0x0002/info) - Inserted: PD 03(e0x20/s3) Inf=
o:=0A>> enclPd=3D20, scsiType=3D0, portMap=3D07, sasAddr=3D5000c50053c57eb5=
,**=0A>> 0000000000000000=0A>> mfi0: 6059 (boot + 19s/0x0002/info) - Insert=
ed: PD 04(e0x20/s4)=0A>> mfi0: 6060 (boot + 19s/0x0002/info) - Inserted: PD=
 04(e0x20/s4) Info:=0A>> enclPd=3D20, scsiType=3D0, portMap=3D00, sasAddr=
=3D5000c50053d364c5,**=0A>> 0000000000000000=0A>> mfi0: 6061 (boot + 19s/0x=
0002/info) - Inserted: PD 05(e0x20/s5)=0A>> mfi0: 6062 (boot + 19s/0x0002/i=
nfo) - Inserted: PD 05(e0x20/s5) Info:=0A>> enclPd=3D20, scsiType=3D0, port=
Map=3D03, sasAddr=3D5000c50053da2b41,**=0A>> 0000000000000000=0A>> mfi0: 60=
63 (boot + 19s/0x0002/info) - Inserted: PD 06(e0x20/s6)=0A>> mfi0: 6064 (bo=
ot + 19s/0x0002/info) - Inserted: PD 06(e0x20/s6) Info:=0A>> enclPd=3D20, s=
csiType=3D0, portMap=3D02, sasAddr=3D4433221102000000,**=0A>> 0000000000000=
000=0A>> mfi0: 6065 (boot + 19s/0x0002/WARN) - PD 06(e0x20/s6) is not a=0A>=
certified=0A>> drive=0A>> mfi0: 6066 (boot + 19s/0x0002/info) - Inserted: P=
D 07(e0x20/s7)=0A>> mfi0: 6067 (boot + 19s/0x0002/info) - Inserted: PD 07(e=
0x20/s7) Info:=0A>> enclPd=3D20, scsiType=3D0, portMap=3D01, sasAddr=3D4433=
221103000000,**=0A>> 0000000000000000=0A>> mfi0: 6068 (boot + 19s/0x0002/WA=
RN) - PD 07(e0x20/s7) is not a=0A>certified=0A>> drive=0A>> mfi0: 6069 (417=
120312s/0x0020/info) - Time established as 03/20/13=0A>> 18:45:12; (32 seco=
nds since power on)=0A>> mfi0: 6070 (417120445s/0x0020/info) - Host driver =
is loaded and=0A>operational=0A>> Timecounters tick every 1.000 msec=0A>> u=
sbus0: 480Mbps High Speed USB v2.0=0A>> usbus1: 480Mbps High Speed USB v2.0=
=0A>> mfid0 on mfi0=0A>> mfid0: 285568MB (584843264 sectors) RAID volume 'B=
ASE' is optimal=0A>> mfid1 on mfi0=0A>> mfid1: 121472MB (248774656 sectors)=
 RAID volume 'L2ARC0' is optimal=0A>> mfid2 on mfi0=0A>> mfid2: 121472MB (2=
48774656 sectors) RAID volume 'ZIL0' is optimal=0A>> mfid3 on mfi0=0A>> mfi=
d3: 1143552MB (2341994496 sectors) RAID volume 'DATASTRIPE1' is=0A>optimal=
=0A>> mfid4 on mfi0=0A>> mfid4: 1143552MB (2341994496 sectors) RAID volume =
'DATASTRIPE2' is=0A>optimal=0A>> ugen0.1: <Intel> at usbus0=0A>> uhub0: <In=
tel EHCI root HUB, class 9/0, rev 2.00/1.00, addr 1> on=0A>usbus0=0A>> ugen=
1.1: <Intel> at usbus1=0A>> uhub1: <Intel EHCI root HUB, class 9/0, rev 2.0=
0/1.00, addr 1> on=0A>usbus1=0A>> uhub0: 2 ports with 2 removable, self pow=
ered=0A>> uhub1: 2 ports with 2 removable, self powered=0A>> ugen0.2: <vend=
or 0x8087> at usbus0=0A>> uhub2: <vendor 0x8087 product 0x0024, class 9/0, =
rev 2.00/0.00, addr=0A>2> on=0A>> usbus0=0A>> ugen1.2: <vendor 0x8087> at u=
sbus1=0A>> uhub3: <vendor 0x8087 product 0x0024, class 9/0, rev 2.00/0.00, =
addr=0A>2> on=0A>> usbus1=0A>> uhub2: 6 ports with 6 removable, self powere=
d=0A>> uhub3: 8 ports with 8 removable, self powered=0A>> ugen0.3: <no manu=
facturer> at usbus0=0A>> uhub4: <no manufacturer Gadget USB HUB, class 9/0,=
 rev 2.00/0.00,=0A>addr 3>=0A>> on usbus0=0A>> SMP: AP CPU #1 Launched!=0A>=
> SMP: AP CPU #7 Launched!=0A>> SMP: AP CPU #3 Launched!=0A>> SMP: AP CPU #=
6 Launched!=0A>> SMP: AP CPU #2 Launched!=0A>> SMP: AP CPU #5 Launched!=0A>=
> SMP: AP CPU #4 Launched!=0A>> Timecounter "TSC-low" frequency 9765827 Hz =
quality 1000=0A>> Root mount waiting for: usbus0=0A>> uhub4: 6 ports with 6=
 removable, self powered=0A>> ugen0.4: <Avocent> at usbus0=0A>> ukbd0: <Key=
board> on usbus0=0A>> kbd0 at ukbd0=0A>> Root mount waiting for: usbus0=0A>=
> Root mount waiting for: usbus0=0A>> ugen0.5: <Avocent> at usbus0=0A>> uma=
ss0: <SCSI Transparent Interface 0> on usbus0=0A>> Trying to mount root fro=
m ufs:/dev/mfid0p3 [rw]...=0A>> cd0 at umass-sim0 bus 0 scbus1 target 0 lun=
 0=0A>> cd0: <iDRAC Virtual CD 0329> Removable CD-ROM SCSI-0 device=0A>> cd=
0: 40.000MB/s transfers=0A>> cd0: Attempt to query device size failed: NOT =
READY, Medium not=0A>present=0A>> da0 at umass-sim0 bus 0 scbus1 target 0 l=
un 1=0A>> da0: <iDRAC Virtual Floppy 0329> Removable Direct Access SCSI-0=
=0A>device=0A>> da0: 40.000MB/s transfers=0A>> da0: Attempt to query device=
 size failed: NOT READY, Medium not=0A>present=0A>> ZFS filesystem version =
5=0A>> ZFS storage pool version 28=0A>> ums0: <Mouse> on usbus0=0A>> ums0: =
3 buttons and [Z] coordinates ID=3D0=0A>> ums1: <Mouse REL> on usbus0=0A>> =
ums1: 3 buttons and [XYZ] coordinates ID=3D0=0A>> igb0: link state changed =
to UP=0A>>=0A>> With this loader.conf:=0A>> vm.kmem_size_max=3D"64G"=0A>> v=
m.kmem_size=3D"48G"=0A>> vfs.zfs.arc_min=3D"4096M"=0A>> vfs.zfs.arc_max=3D"=
15872M"=0A>> vfs.zfs.txg.synctime=3D"1"=0A>> vfs.zfs.txg.timeout=3D"5"=0A>>=
 vfs.zfs.write_limit_override=3D"**0"=0A>> vfs.zfs.txg.write_limit_**overri=
de=3D"0"=0A>> vfs.zfs.zil_disable=3D"0"=0A>>=0A>> And with this sysctl.conf=
:=0A>> security.bsd.see_other_uids=3D0=0A>> net.inet.tcp.blackhole=3D2=0A>>=
 net.inet.udp.blackhole=3D1=0A>> net.inet.tcp.sendspace=3D65536=0A>> net.in=
et.tcp.recvspace=3D65536=0A>> net.link.ether.inet.log_arp_**wrong_iface=3D0=
=0A>> kern.ipc.somaxconn=3D32768=0A>> kern.threads.max_threads_per_**proc=
=3D16384=0A>> kern.maxfiles=3D262144=0A>> kern.maxfilesperproc=3D131072=0A>=
> kern.ipc.nmbclusters=3D65536=0A>> kern.corefile=3D"/var/coredumps/**%U.%N=
.%P.core"=0A>>=0A>> vfs.zfs.txg.timeout=3D5=0A>> vfs.zfs.l2arc_headroom=3D1=
6=0A>> vfs.zfs.l2arc_write_max=3D**26843546=0A>> vfs.zfs.l2arc_write_boost=
=3D**26843546=0A>> vfs.zfs.vdev.min_pending=3D4=0A>> vfs.zfs.vdev.max_pendi=
ng=3D64=0A>> vfs.zfs.l2arc_noprefetch=3D1=0A>> vfs.zfs.dedup.prefetch=3D0=
=0A>> vfs.zfs.prefetch_disable=3D1=0A>>=0A>> I will use this server as a my=
sql-5.6 dbserver so I have a root=0A>partition=0A>> using a hw raid1 and a =
/DATAZFS partition, follows configuration:=0A>>=0A>> root# mfiutil show dri=
ves=0A>> mfi0 Physical Drives:=0A>>=C2=A0 0 (=C2=A0 279G) ONLINE <SEAGATE S=
T3300657SS ES65> SAS E1:S0=0A>>=C2=A0 1 (=C2=A0 279G) ONLINE <SEAGATE ST330=
0657SS ES65> SAS E1:S1=0A>>=C2=A0 2 (=C2=A0 558G) ONLINE <SEAGATE ST3600057=
SS ES65> SAS E1:S2=0A>>=C2=A0 3 (=C2=A0 558G) ONLINE <SEAGATE ST3600057SS E=
S65> SAS E1:S3=0A>>=C2=A0 4 (=C2=A0 558G) ONLINE <SEAGATE ST3600057SS ES65>=
 SAS E1:S4=0A>>=C2=A0 5 (=C2=A0 558G) ONLINE <SEAGATE ST3600057SS ES65> SAS=
 E1:S5=0A>>=C2=A0 6 (=C2=A0 119G) ONLINE <OCZ-VERTEX4 1.4> SATA E1:S6=0A>>=
=C2=A0 7 (=C2=A0 119G) ONLINE <OCZ-VERTEX4 1.4> SATA E1:S7=0A>> root# mfiut=
il show volumes=0A>> mfi0 Volumes:=0A>>=C2=A0  Id=C2=A0 =C2=A0  Size=C2=A0 =
=C2=A0 Level=C2=A0  Stripe=C2=A0 State=C2=A0  Cache=C2=A0  Name=0A>>=C2=A0 =
mfid0 (=C2=A0 278G) RAID-1=C2=A0 =C2=A0 =C2=A0 64k OPTIMAL Disabled <BASE>=
=0A>>=C2=A0 mfid1 (=C2=A0 118G) RAID-0=C2=A0 =C2=A0 =C2=A0 64k OPTIMAL Disa=
bled <L2ARC0>=0A>>=C2=A0 mfid2 (=C2=A0 118G) RAID-0=C2=A0 =C2=A0 =C2=A0 64k=
 OPTIMAL Disabled <ZIL0>=0A>>=C2=A0 mfid3 ( 1116G) RAID-1=C2=A0 =C2=A0 =C2=
=A0 64k OPTIMAL Disabled <DATA>=0A>>=0A>> I have a /DATAZFS partition, too:=
=0A>> root# zfs get all DATAZFS=0A>> NAME=C2=A0 =C2=A0  PROPERTY=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 VALUE=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 SOURCE=0A>> DATAZFS=C2=A0 type=C2=A0 =C2=A0=
 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 filesystem=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0  -=0A>> DATAZFS=C2=A0 creation=C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 Thu Mar 21 10:45 2013=C2=A0 -=0A>> DATAZFS=
=C2=A0 used=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 5=
4.2G=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> D=
ATAZFS=C2=A0 available=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  1.02T=C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=
=C2=A0 referenced=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 54.2G=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 c=
ompressratio=C2=A0 =C2=A0 =C2=A0 =C2=A0  1.00x=C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 mounted=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  yes=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 quota=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  none=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 reservat=
ion=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  none=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 recordsize=C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 16K=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 local=0A>> DATAZFS=C2=A0 mountpoint=
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 /DATAZFS=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 sharenfs=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 off=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0 checksum=C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 on=C2=A0 =C2=A0 =C2=A0 =C2=A0=
 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 comp=
ression=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  off=C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0 atime=
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  off=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 local=0A>> DATAZFS=
=C2=A0 devices=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  on=C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>=
> DATAZFS=C2=A0 exec=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=
 =C2=A0 on=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0  default=0A>> DATAZFS=C2=A0 setuid=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=
 =C2=A0 =C2=A0 =C2=A0 on=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 readonly=C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 off=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0 jailed=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 off=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0 sn=
apdir=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  hidden=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 aclmo=
de=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  discard=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0 aclinh=
erit=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 restricted=C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 canmount=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 on=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=C2=A0 xattr=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  off=C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 temporary=0A>> DATAZFS=C2=A0 =
copies=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 1=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=
=0A>> DATAZFS=C2=A0 version=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=
  5=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 -=0A>> DATAZFS=C2=A0 utf8only=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 off=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 normalization=C2=A0 =C2=A0 =C2=A0 =C2=A0  n=
one=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  -=0A>> D=
ATAZFS=C2=A0 casesensitivity=C2=A0 =C2=A0 =C2=A0  sensitive=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 vscan=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  off=C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0 nbmand=
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 off=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZF=
S=C2=A0 sharesmb=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 off=C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>>=
 DATAZFS=C2=A0 refquota=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 non=
e=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A=
>> DATAZFS=C2=A0 refreservation=C2=A0 =C2=A0 =C2=A0 =C2=A0 none=C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  default=0A>> DATAZFS=
=C2=A0 primarycache=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 all=C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=
=C2=A0 secondarycache=C2=A0 =C2=A0 =C2=A0 =C2=A0 all=C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=C2=A0=
 usedbysnapshots=C2=A0 =C2=A0 =C2=A0  0=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 usedbydatase=
t=C2=A0 =C2=A0 =C2=A0 =C2=A0  54.2G=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=C2=A0 usedbychildren=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 1.41M=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 -=0A>> DATAZFS=C2=A0 usedbyrefreservation=C2=A0 0=C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=
=C2=A0 logbias=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  latency=C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 default=0A>> DATAZFS=
=C2=A0 dedup=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  off=C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 local=0A=
>> DATAZFS=C2=A0 mlslabel=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0  -=0A>> DATAZFS=C2=A0 sync=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 standard=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
 default=0A>> DATAZFS=C2=A0 refcompressratio=C2=A0 =C2=A0 =C2=A0 1.00x=C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>> DATAZFS=
=C2=A0 written=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  54.2G=C2=A0=
 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 -=0A>>=0A>> root# =
zpool status -v=0A>>=C2=A0  pool: DATAZFS=0A>>=C2=A0 state: ONLINE=0A>>=C2=
=A0  scan: none requested=0A>> config:=0A>>=0A>> NAME=C2=A0 =C2=A0 =C2=A0 =
=C2=A0 STATE=C2=A0 =C2=A0  READ WRITE CKSUM=0A>> DATAZFS=C2=A0 =C2=A0  ONLI=
NE=C2=A0 =C2=A0 =C2=A0  0=C2=A0 =C2=A0  0=C2=A0 =C2=A0  0=0A>>=C2=A0  mfid3=
=C2=A0 =C2=A0  ONLINE=C2=A0 =C2=A0 =C2=A0  0=C2=A0 =C2=A0  0=C2=A0 =C2=A0  =
0=0A>> logs=0A>>=C2=A0  mfid2=C2=A0 =C2=A0  ONLINE=C2=A0 =C2=A0 =C2=A0  0=
=C2=A0 =C2=A0  0=C2=A0 =C2=A0  0=0A>> cache=0A>>=C2=A0  mfid1=C2=A0 =C2=A0 =
 ONLINE=C2=A0 =C2=A0 =C2=A0  0=C2=A0 =C2=A0  0=C2=A0 =C2=A0  0=0A>>=0A>> er=
rors: No known data errors=0A>>=0A>> root# mfiutil show firmware=0A>> mfi0 =
Firmware Package Version: 21.1.0-0007=0A>> mfi0 Firmware Images:=0A>> Name=
=C2=A0 Version=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 Date=C2=A0 =C2=A0 =C2=A0 =C2=A0  Time=C2=A0 =C2=A0=
 =C2=A0 =C2=A0  Status=0A>> BIOS=C2=A0 5.31.01_4.12.05.00_0x05110000=C2=A0 =
 6/20/2012=0A>>=C2=A0 =C2=A0 6/20/2012=0A>>=C2=A0  active=0A>> CTLR=C2=A0 4=
.00-0014=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 Aug 04 2011=C2=A0 12:49:17=C2=A0 =C2=A0  active=0A>> PCLI=C2=A0 =
05.00-03:#%00008=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  Feb 17 20=
11=C2=A0 14:03:12=C2=A0 =C2=A0  active=0A>> APP=C2=A0  3.130.05-1796=C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 Jul 27 2012=C2=A0 1=
5:58:31=C2=A0 =C2=A0  active=0A>> NVDT=C2=A0 2.1108.03-0094=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0  Jun 22 2012=C2=A0 12:40:55=C2=A0=
 =C2=A0  active=0A>> BTBL=C2=A0 2.03.00.00-0004=C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 Jul 27 2012=C2=A0 16:39:29=C2=A0 =C2=A0  active=
=0A>> BOOT=C2=A0 06.253.57.219=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 9/9/2010=C2=A0 =C2=A0  15:32:25=C2=A0 =C2=A0  active=0A>>=
 root#=0A>>=0A>> Then, I installed CentOS on the same hardware, same versio=
n of mysql=0A>> (5.6.10-enterprise and 5.6.10-GA), and I tried an OLTP test=
 (included=0A>in=0A>> sysbench-0.5) executing 2000000 queries on 10000000 r=
ows, with these=0A>> results (transactions per seconds):=0A>>=0A>> BSD 9.1 =
amd64 ZFS (without tweaks):=0A>> 1 thread: 2780=0A>> 2 threads: 4371=0A>> 4=
 threads: 6876=0A>> 8 threads: 8202=0A>> 12 threads: 8077=0A>> 16 threads: =
7780=0A>> 24 threads: 7563=0A>> 32 threads: 7632=0A>> 48 threads: 7960=0A>>=
 64 threads: 8062=0A>>=0A>> BSD 9.1 amd64 ZFS (with tweaks):=0A>> 1 thread:=
 2589=0A>> 2 threads: 4679=0A>> 4 threads: 6438=0A>> 8 threads: 7073=0A>> 1=
2 threads: 7121=0A>> 16 threads: 7227=0A>> 24 threads: 7132=0A>> 32 threads=
: 7273=0A>> 48 threads: 7623=0A>> 64 threads: 7672=0A>>=0A>> CentOS 6.3 (ex=
t4, without tweaks):=0A>> 1 thread: 4163=0A>> 2 threads: 7653=0A>> 4 thread=
s: 10905=0A>> 8 threads: 12511=0A>> 12 threads: 13556=0A>> 16 threads: 1483=
2=0A>> 24 threads: 16270=0A>> 32 threads: 16733=0A>> 48 threads: 16925=0A>>=
 64 threads: 16895=0A>>=0A>>=0A>> Well, any of you could ask "why zfs, plea=
se try ufs that is better":=0A>you=0A>> are right, but changing ZFS with UF=
S doesn't change so much:=0A>>=0A>http://lists.freebsd.org/**pipermail/free=
bsd-fs/2013-**March/016747.html<http://lists.freebsd.org/pipermail/freebsd-=
fs/2013-March/016747.html>=0A>>=0A>> So I hope I'm missing something becaus=
e the performances of a=0A>standard=0A>> CentOS installation on the same ha=
rdware performs much better than a=0A>9.1=0A>> amd64 installation (I'm usin=
g the standard ULE scheduler included in=0A>the=0A>> 9.1 release).=0A>>=0A>=
> Could anyone help me?=0A>>=0A>> Thanks,=0A>> d.=0A>> ____________________=
__________**_________________=0A>> freebsd-performance@freebsd.**org=0A><fr=
eebsd-performance@freebsd.org>mailing list=0A>>=0A>http://lists.freebsd.org=
/**mailman/listinfo/freebsd-**performance<http://lists.freebsd.org/mailman/=
listinfo/freebsd-performance>=0A>> To unsubscribe, send any mail to "freebs=
d-performance-**=0A>> unsubscribe@freebsd.org=0A><freebsd-performance-unsub=
scribe@freebsd.org>"=0A>>=0A=0A--=0Ad.=0A__________________________________=
_____________=0Afreebsd-performance@freebsd.org mailing list=0Ahttp://lists=
.freebsd.org/mailman/listinfo/freebsd-performance=0ATo unsubscribe, send an=
y mail to "freebsd-performance-unsubscribe@freebsd.org"
From owner-freebsd-performance@FreeBSD.ORG  Sat Mar 23 07:44:34 2013
Return-Path: <owner-freebsd-performance@FreeBSD.ORG>
Delivered-To: freebsd-performance@freebsd.org
Received: from mx1.freebsd.org (mx1.freebsd.org
 [IPv6:2001:1900:2254:206a::19:1])
 by hub.freebsd.org (Postfix) with ESMTP id 8632261A
 for <freebsd-performance@freebsd.org>; Sat, 23 Mar 2013 07:44:34 +0000 (UTC)
 (envelope-from davide.damico@contactlab.com)
Received: from mail2.shared.smtp.contactlab.it
 (mail2.shared.smtp.contactlab.it [93.94.37.7])
 by mx1.freebsd.org (Postfix) with ESMTP id 14A42979
 for <freebsd-performance@freebsd.org>; Sat, 23 Mar 2013 07:44:33 +0000 (UTC)
DKIM-Signature: v=1; a=rsa-sha256; d=contactlab.it; s=clab1; c=relaxed/relaxed;
 q=dns/txt; i=@contactlab.it; t=1364024671;
 h=From:Subject:Date:To:MIME-Version:Content-Type;
 bh=sVP3v1AW07tk+cpo0iIlPaC3P7T3DOUwyUlqA589++k=;
 b=VHxXWhW9/BZ619E3u+b6/OD4A5E8O2kPCaU1KKKVQT3jY5LxuovJ3eH4cFQ6kAbO
 Ina/P4HYUs7OmJlTl0tBWAgaTU6TqFO6+aUOZ3arVqlGrqGNv8pA5ZsX/SDPEGnc
 OsGhAzw078HrPOFG+i7ptFGYWzEzQ4vWwt5yWoFyj+E=;
Received: from [213.92.90.12] ([213.92.90.12:26194] helo=mail3.tomato.it)
 by t.contactlab.it (envelope-from <davide.damico@contactlab.com>)
 (ecelerity 3.5.1.37854 r(Momo-dev:3.5.1.0)) with ESMTP
 id FB/91-24145-F5D5D415; Sat, 23 Mar 2013 08:44:31 +0100
Received: from mx3-master.housing.tomato.lan ([172.16.7.55])
 by mail3.tomato.it with smtp (Exim 4.80.1 (FreeBSD))
 (envelope-from <davide.damico@contactlab.com>) id 1UJJ7v-000Pn0-CO
 for freebsd-performance@freebsd.org; Sat, 23 Mar 2013 08:44:31 +0100
Received: (qmail 99132 invoked by uid 80); 23 Mar 2013 07:44:31 -0000
To: Paul Pathiakis <pathiaki2@yahoo.com>
Subject: Re: FreeBSD 9.1 vs CentOS 6.3
X-PHP-Script: uebmeil.sys.tomatointeractive.it/index.php for 172.16.16.227
X-PHP-Originating-Script: 0:main.inc
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8;
 format=flowed
Content-Transfer-Encoding: 7bit
Date: Sat, 23 Mar 2013 08:44:31 +0100
From: Davide D'Amico <davide.damico@contactlab.com>
Organization: ContactLab
Mail-Reply-To: <davide.damico@contactlab.com>
In-Reply-To: <1363998883.22604.YahooMailNeo@web141401.mail.bf1.yahoo.com>
References: <CAF3xD3nDDUYB94TS9AUUQ=CPztB1S8mU4fRwNB5GupmK8MgwXg@mail.gmail.com>
 <ebfc033f-d8db-4162-86be-db013e3d7eb2@email.android.com>
 <1363998883.22604.YahooMailNeo@web141401.mail.bf1.yahoo.com>
Message-ID: <5d10e2a0b0f6477293459a26df1fc272@sys.tomatointeractive.it>
X-Sender: davide.damico@contactlab.com
User-Agent: Roundcube Webmail/0.8.5
Cc: freebsd-performance@freebsd.org,
 =?UTF-8?Q?=D0=95=D0=B2=D0=B3=D0=B5=D0=BD=D0=B8=D0=B9_=D0=A5=D0=BE?=
 =?UTF-8?Q?=D1=80=D0=BA=D0=B8=D0=BD?= <khorkin.eugeniy@gmail.com>
X-BeenThere: freebsd-performance@freebsd.org
X-Mailman-Version: 2.1.14
Precedence: list
Reply-To: davide.damico@contactlab.com
List-Id: Performance/tuning <freebsd-performance.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/options/freebsd-performance>, 
 <mailto:freebsd-performance-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-performance>;
List-Post: <mailto:freebsd-performance@freebsd.org>
List-Help: <mailto:freebsd-performance-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-performance>, 
 <mailto:freebsd-performance-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Sat, 23 Mar 2013 07:44:34 -0000

Il 23.03.2013 01:34 Paul Pathiakis ha scritto:
> Hi,
> 
> There are several things about this that are highly suspect.
> 
> First, wipe out the hardware RAID. The processor doing RAID
> computation is, probably, MUCH slower than a core on the CPU. Even if
> it's RAID-1 (Simple Mirror) this RAID card is performing tasks that is
> does not need to do including replicating writes to two targets from
> the controller or checking it's cache, battery, etc. If it's possible
> to disable the onboard cache, do it.

Hi Paul,
thanks for your suggestions (some of them I've applied before starting 
any consideration, like disabling all on-disk caches or controller 
buffers) I'll try next monday.

Anyway, the fact is that using the same hardware configuration 
(raid1+raid10) I saw that a centos 6.x outperformed freebsd 9.1.
Another test I made yesterday was: on the same hardware I installed 
vmware esx 5.x and created a vm with centos inside it. The result was 
really impressive: the centos vm outperformed the 'real' freebsd 9.1 too 
and checking vmware performances graphs I didn't see any huge need for a 
massive throughput (I saw values from KBps to 10MBps), instead I saw a 
big use of CPU (using OLTP tests with a concurrency of 32 threads it's 
performaces began to slow down).

I don't know is using some magic value for HZ or setting some trick 
with scheduler, I could gain something: I hope so, because I don't want 
to "pinguinate" my farm :)

Thanks,
d.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1363998883.22604.YahooMailNeo>