Skip to contentSkip to conversation

3Ware 9500S-12 RAID controller -- poor performance

DISCUSSION subbie subbie 5 participants 11 messages Sep 30, 2005 – Oct 6, 2005

lore
Share & tools
lore
#0 lore raw

Dear list,

After almost two weeks of experimentation, google searches and reading of posts, bug reports and discussions I'm still far from an answer. I'm hoping someone on this list could shed some light on the subject.

I'm using a 3Ware 9500S-12 card and am able to produce up to 400MB/s sustained read from my 12-disk 4.1TB RAID5 SATA array, 128MB cache onboard, ext3 formatted.
All is well when performing a single read -- it
works nice and fast.

The system is a web server, serving mid-size files (50MB, each, on average). All hell breaks loose when doing many concurrent reads, anywhere between 200 to 400 concurrent streams things simply grind to a halt and the system transfers a maximum of 12-14MB/s.

I'm in the process of clearing up the array (this would take some time) and restructuring it to JBOD mode in order to use each disk individually. I will use a filesystem more suitable to streaming large files, such as XFS. But this would take time and I would very much appreciate the advice of people in the
know if this is going to help at all. It's hard for
me to make extreme experimentation (deleting, formatting, reformatting) as this is a productio n system with many files that I have no other place to dump until they can be safely removed. Though I'm working on dumping them slowly to other, remote, machines.

I'm running the latest kernel, 2.6.13.2 and the latest 3Ware driver, taken from the 3ware.com web site which upon insmod, updates the card's firmware to the latest version as well.

In my experiments, I've tried using larger readahead, currently at 16k (this helps, higher values do not seem to help much), using the deadline scheduler for this device, booting the system with the 'noapic' option and playing with a bunch of VM tunable parameters which I'm not sure that I should really be touching. At the moment only the readahead modification is used as the other stuff simply didn't help at all.

With the stock kernel shipped with my distribution, 2.6.8 and its old 3ware driver things were just as
worse but manifested themselves differently. The
system was visibly (top, vmstat...) spending most of its time in io-wait and load average was extremely
high, in the area of 10 to 20. With the recent
kernel and driver mentioned above, the excessive io-wait and load seems to have been resolved and observed loadavg is between 1 and 4.

I don't have much experience with systems that are supposed to stream many files concurrently off a hardware RAID of this configuration, but my gut feeling is that something is very wrong and I should be seeing a much higher read throughput.

Trying to preempt people's questions I've tried to include as much information as possible, a lot of stuff is pasted below.

I've just seen that the 3ware driver shares the same IRQ with my ethernet card, which has got me a little worried, should I be?

System uptime, exactly 1 day:

# cat /proc/interrupts 
           CPU0       CPU1       
  0:   21619638          0          XT-PIC  timer
  2:          0          0          XT-PIC  cascade
  8:          4          0          XT-PIC  rtc
 10:  268753224          0          XT-PIC  3w-9xxx,
eth0
 14:         11          0          XT-PIC  ide0
 15:     337881          0          XT-PIC  libata
NMI:          0          0 
LOC:   21110402   21557685 
ERR:          0
MIS:          0
# free
            total       used       free     shared   
buffers     cached
Mem:       2075260    2024724      50536          0   
   5184    1388408
-/+ buffers/cache:     631132    1444128
Swap:      3903784          0    3903784
# vmstat -n 1  (output of the last few seconds):
procs -----------memory---------- ---swap--
-----io---- --system-- ----cpu----
 0  0      0  49932   4760 1392980    0    0 15636   
32 3169  3697  4  6 30 60
 0  0      0  50816   4752 1392376    0    0  5844    
0 3114  3929  3  5 91  1
 0  0      0  50924   4772 1391404    0    0  9360    
0 3187  4348  6  6 76 13
 0  2      0  50552   4780 1391532    0    0 24976   
44 4077  3906  3  7 65 25
 0  1      0  50444   4780 1392688    0    0 20192    
0 5048  3914  7  8 56 30
 0  1      0  50568   4756 1392508    0    0 21248    
0 4060  3603  4  6 48 41
 0  0      0  50704   4724 1392268    0    0 30004    
0 3834  3369  4  9 65 22
 0  3      0  50556   4728 1392468    0    0  3248 
1832 2974  4514  2  5 58 35
 0  3      0  50308   4724 1392200    0    0  1288  
336 1766  1886  1  3 50 47
 0  4      0  50308   4732 1391852    0    0  2300  
408 1919  2158  0  3 51 46
 0  4      0  50556   4736 1390692    0    0  1856  
532 1488  1846  3  1 50 46
 0  3      0  50680   4740 1390620    0    0  4016 
1296 1577  1682  2  2 50 47
 0  3      0  50432   4752 1391628    0    0  2180   
72 1730  1945  2  2 51 46
 2  2      0  49924   4772 1391540    0    0 44372  
564 3403  2847  4  5 50 42
 0  0      0  50684   4784 1391528    0    0 28640  
216 3804  3847  7  8 69 16 
# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      : Intel(R) Xeon(TM) CPU 3.00GHz
stepping        : 1
cpu MHz         : 2993.035
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 1
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8
apic sep mtrr pge mca cmov 
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm
pbe nx lm pni monitor ds_c
pl cid cx16 xtpr
bogomips        : 5993.68

processor : 1
vendor_id : GenuineIntel
cpu family : 15
model : 4
model name : Intel(R) Xeon(TM) CPU 3.00GHz
stepping : 1
cpu MHz : 2993.035
cache size : 1024 KB
physical id : 3
siblings : 2
core id : 3
cpu cores : 1
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 5
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8
apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm pni monitor ds_c
pl cid cx16 xtpr
bogomips : 5985.52

# lspci
0000:00:00.0 Host bridge: Intel Corporation E7520
Memory Controller Hub (rev 0c)
0000:00:00.1 ff00: Intel Corporation E7525/E7520 Error
Reporting Registers (rev 
0c)
0000:00:01.0 System peripheral: Intel Corporation
E7520 DMA Controller (rev 0c)
0000:00:02.0 PCI bridge: Intel Corporation
E7525/E7520/E7320 PCI Express Port A 
(rev 0c)
0000:00:04.0 PCI bridge: Intel Corporation E7525/E7520
PCI Express Port B (rev 0
c)
0000:00:05.0 PCI bridge: Intel Corporation E7520 PCI
Express Port B1 (rev 0c)
0000:00:06.0 PCI bridge: Intel Corporation E7520 PCI
Express Port C (rev 0c)
0000:00:1d.0 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB UHCI 
Controller #1 (rev 02)
0000:00:1d.1 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB UHCI 
Controller #2 (rev 02)
0000:00:1d.2 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB UHCI 
Controller #3 (rev 02)
0000:00:1d.7 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB2 EHCI
 Controller (rev 02)
0000:00:1e.0 PCI bridge: Intel Corporation 82801 PCI
Bridge (rev c2)
0000:00:1f.0 ISA bridge: Intel Corporation 82801EB/ER
(ICH5/ICH5R) LPC Interface
 Bridge (rev 02)
0000:00:1f.1 IDE interface: Intel Corporation
82801EB/ER (ICH5/ICH5R) IDE Contro
ller (rev 02)
0000:00:1f.2 IDE interface: Intel Corporation 82801EB
(ICH5) SATA Controller (re
v 02)
0000:00:1f.3 SMBus: Intel Corporation 82801EB/ER
(ICH5/ICH5R) SMBus Controller (
rev 02)
0000:01:00.0 PCI bridge: Intel Corporation 6700PXH PCI
Express-to-PCI Bridge A (
rev 09)
0000:01:00.1 PIC: Intel Corporation 6700/6702PXH
I/OxAPIC Interrupt Controller A
 (rev 09)
0000:01:00.2 PCI bridge: Intel Corporation 6700PXH PCI
Express-to-PCI Bridge B (
rev 09)
0000:01:00.3 PIC: Intel Corporation 6700PXH I/OxAPIC
Interrupt Controller B (rev
 09)
0000:03:01.0 RAID bus controller: 3ware Inc
9xxx-series SATA-RAID
0000:05:00.0 Ethernet controller: Marvell Technology
Group Ltd. 88E8050 Gigabit
Ethernet Controller (rev 18)
0000:07:04.0 Ethernet controller: Intel Corporation
82541GI/PI Gigabit Ethernet 
Controller (rev 05)
0000:07:0c.0 VGA compatible controller: ATI
Technologies Inc Rage XL (rev 27)

As you can see, this is a fairly recent motherboard that's supposed to perform well, though I don't know the manufacturer of this board as the machine is hosted and I don't have physical access, though I can ask them if anyone would like to know.

The ethernet card actually being used is the Intel E1000 with NAPI support compiled.

If there's any bit of information that's missing, please let me know and I'd he happy to provide it quickly.

If you can suggest a better (non Netapp, EMC, etc) solution that is somehow affordable and can provide the very high read throughputs please let me know, I'm very interested in solutions that can staturate multiple gigabit links (of course, using more than one machine ;)

Please CC me on any replies as I'm not subscribed to the list.

Thank you for listening!

__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com

#1 lore raw

Why are you booting with 'noapic'.. in my experience that will seriously impact interrupt performance. Use the APIC if you've got it, which in this case you definitely do.

Yes, having your gigabit NIC and RAID controller on the same IRQ (in PIC mode) could definitely me a source of trouble.

In your web server testing, were you using an external traffic generator or an on-host process? If you try on-host (eliminating the network throughput and related interrupts) does performance improve?

So two biggest suggestions:

- Use the APIC. It is your friend.

- It looks like the 3ware card and gigabit nic are on different busses, but the pirq lines are being routed to the same legacy interrupt in PIC mode. So APIC mode should avoid that problem. If the controller and nic are actually on the same bus, separate them.

Regards,
Ian Morgan

Signature · 5 lines
-------------------------------------------------------------------
 Ian E. Morgan          Vice President & C.O.O.       Webcon, Inc.
 imorgan at webcon dot ca        PGP: #2DA40D07      www.webcon.ca
    *  Customized Linux network solutions for your business  *
-------------------------------------------------------------------
Quoted 277 lines · subbie subbie

Dear list,

After almost two weeks of experimentation, google searches and reading of posts, bug reports and discussions I'm still far from an answer. I'm hoping someone on this list could shed some light on the subject.

I'm using a 3Ware 9500S-12 card and am able to produce up to 400MB/s sustained read from my 12-disk 4.1TB RAID5 SATA array, 128MB cache onboard, ext3 formatted.
All is well when performing a single read -- it
works nice and fast.

The system is a web server, serving mid-size files (50MB, each, on average). All hell breaks loose when doing many concurrent reads, anywhere between 200 to 400 concurrent streams things simply grind to a halt and the system transfers a maximum of 12-14MB/s.

I'm in the process of clearing up the array (this would take some time) and restructuring it to JBOD mode in order to use each disk individually. I will use a filesystem more suitable to streaming large files, such as XFS. But this would take time and I would very much appreciate the advice of people in the
know if this is going to help at all. It's hard for
me to make extreme experimentation (deleting, formatting, reformatting) as this is a productio n system with many files that I have no other place to dump until they can be safely removed. Though I'm working on dumping them slowly to other, remote, machines.

I'm running the latest kernel, 2.6.13.2 and the latest 3Ware driver, taken from the 3ware.com web site which upon insmod, updates the card's firmware to the latest version as well.

In my experiments, I've tried using larger readahead, currently at 16k (this helps, higher values do not seem to help much), using the deadline scheduler for this device, booting the system with the 'noapic' option and playing with a bunch of VM tunable parameters which I'm not sure that I should really be touching. At the moment only the readahead modification is used as the other stuff simply didn't help at all.

With the stock kernel shipped with my distribution, 2.6.8 and its old 3ware driver things were just as
worse but manifested themselves differently. The
system was visibly (top, vmstat...) spending most of its time in io-wait and load average was extremely
high, in the area of 10 to 20. With the recent
kernel and driver mentioned above, the excessive io-wait and load seems to have been resolved and observed loadavg is between 1 and 4.

I don't have much experience with systems that are supposed to stream many files concurrently off a hardware RAID of this configuration, but my gut feeling is that something is very wrong and I should be seeing a much higher read throughput.

Trying to preempt people's questions I've tried to include as much information as possible, a lot of stuff is pasted below.

I've just seen that the 3ware driver shares the same IRQ with my ethernet card, which has got me a little worried, should I be?

System uptime, exactly 1 day:

# cat /proc/interrupts
          CPU0       CPU1
 0:   21619638          0          XT-PIC  timer
 2:          0          0          XT-PIC  cascade
 8:          4          0          XT-PIC  rtc
10:  268753224          0          XT-PIC  3w-9xxx,
eth0
14:         11          0          XT-PIC  ide0
15:     337881          0          XT-PIC  libata
NMI:          0          0
LOC:   21110402   21557685
ERR:          0
MIS:          0
# free
           total       used       free     shared
buffers     cached
Mem:       2075260    2024724      50536          0
  5184    1388408
-/+ buffers/cache:     631132    1444128
Swap:      3903784          0    3903784
# vmstat -n 1  (output of the last few seconds):
procs -----------memory---------- ---swap--
-----io---- --system-- ----cpu----
0  0      0  49932   4760 1392980    0    0 15636
32 3169  3697  4  6 30 60
0  0      0  50816   4752 1392376    0    0  5844
0 3114  3929  3  5 91  1
0  0      0  50924   4772 1391404    0    0  9360
0 3187  4348  6  6 76 13
0  2      0  50552   4780 1391532    0    0 24976
44 4077  3906  3  7 65 25
0  1      0  50444   4780 1392688    0    0 20192
0 5048  3914  7  8 56 30
0  1      0  50568   4756 1392508    0    0 21248
0 4060  3603  4  6 48 41
0  0      0  50704   4724 1392268    0    0 30004
0 3834  3369  4  9 65 22
0  3      0  50556   4728 1392468    0    0  3248
1832 2974  4514  2  5 58 35
0  3      0  50308   4724 1392200    0    0  1288
336 1766  1886  1  3 50 47
0  4      0  50308   4732 1391852    0    0  2300
408 1919  2158  0  3 51 46
0  4      0  50556   4736 1390692    0    0  1856
532 1488  1846  3  1 50 46
0  3      0  50680   4740 1390620    0    0  4016
1296 1577  1682  2  2 50 47
0  3      0  50432   4752 1391628    0    0  2180
72 1730  1945  2  2 51 46
2  2      0  49924   4772 1391540    0    0 44372
564 3403  2847  4  5 50 42
0  0      0  50684   4784 1391528    0    0 28640
216 3804  3847  7  8 69 16
# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      : Intel(R) Xeon(TM) CPU 3.00GHz
stepping        : 1
cpu MHz         : 2993.035
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 1
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 5
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8
apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm
pbe nx lm pni monitor ds_c
pl cid cx16 xtpr
bogomips        : 5993.68

processor : 1
vendor_id : GenuineIntel
cpu family : 15
model : 4
model name : Intel(R) Xeon(TM) CPU 3.00GHz
stepping : 1
cpu MHz : 2993.035
cache size : 1024 KB
physical id : 3
siblings : 2
core id : 3
cpu cores : 1
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 5
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8
apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm pni monitor ds_c
pl cid cx16 xtpr
bogomips : 5985.52

# lspci
0000:00:00.0 Host bridge: Intel Corporation E7520
Memory Controller Hub (rev 0c)
0000:00:00.1 ff00: Intel Corporation E7525/E7520 Error
Reporting Registers (rev
0c)
0000:00:01.0 System peripheral: Intel Corporation
E7520 DMA Controller (rev 0c)
0000:00:02.0 PCI bridge: Intel Corporation
E7525/E7520/E7320 PCI Express Port A
(rev 0c)
0000:00:04.0 PCI bridge: Intel Corporation E7525/E7520
PCI Express Port B (rev 0
c)
0000:00:05.0 PCI bridge: Intel Corporation E7520 PCI
Express Port B1 (rev 0c)
0000:00:06.0 PCI bridge: Intel Corporation E7520 PCI
Express Port C (rev 0c)
0000:00:1d.0 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB UHCI
Controller #1 (rev 02)
0000:00:1d.1 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB UHCI
Controller #2 (rev 02)
0000:00:1d.2 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB UHCI
Controller #3 (rev 02)
0000:00:1d.7 USB Controller: Intel Corporation
82801EB/ER (ICH5/ICH5R) USB2 EHCI
Controller (rev 02)
0000:00:1e.0 PCI bridge: Intel Corporation 82801 PCI
Bridge (rev c2)
0000:00:1f.0 ISA bridge: Intel Corporation 82801EB/ER
(ICH5/ICH5R) LPC Interface
Bridge (rev 02)
0000:00:1f.1 IDE interface: Intel Corporation
82801EB/ER (ICH5/ICH5R) IDE Contro
ller (rev 02)
0000:00:1f.2 IDE interface: Intel Corporation 82801EB
(ICH5) SATA Controller (re
v 02)
0000:00:1f.3 SMBus: Intel Corporation 82801EB/ER
(ICH5/ICH5R) SMBus Controller (
rev 02)
0000:01:00.0 PCI bridge: Intel Corporation 6700PXH PCI
Express-to-PCI Bridge A (
rev 09)
0000:01:00.1 PIC: Intel Corporation 6700/6702PXH
I/OxAPIC Interrupt Controller A
(rev 09)
0000:01:00.2 PCI bridge: Intel Corporation 6700PXH PCI
Express-to-PCI Bridge B (
rev 09)
0000:01:00.3 PIC: Intel Corporation 6700PXH I/OxAPIC
Interrupt Controller B (rev
09)
0000:03:01.0 RAID bus controller: 3ware Inc
9xxx-series SATA-RAID
0000:05:00.0 Ethernet controller: Marvell Technology
Group Ltd. 88E8050 Gigabit
Ethernet Controller (rev 18)
0000:07:04.0 Ethernet controller: Intel Corporation
82541GI/PI Gigabit Ethernet
Controller (rev 05)
0000:07:0c.0 VGA compatible controller: ATI
Technologies Inc Rage XL (rev 27)

As you can see, this is a fairly recent motherboard that's supposed to perform well, though I don't know the manufacturer of this board as the machine is hosted and I don't have physical access, though I can ask them if anyone would like to know.

The ethernet card actually being used is the Intel E1000 with NAPI support compiled.

If there's any bit of information that's missing, please let me know and I'd he happy to provide it quickly.

If you can suggest a better (non Netapp, EMC, etc) solution that is somehow affordable and can provide the very high read throughputs please let me know, I'm very interested in solutions that can staturate multiple gigabit links (of course, using more than one machine ;)

Please CC me on any replies as I'm not subscribed to the list.

Thank you for listening!

#2 lore raw

Dear Ian,

'noapic' was a recommendation by 3Ware / AMCC tech
support. It did not help at all, as expected.
Unfortunately they did not have any other recommendations.

I've now removed 'noapic' and unfortunately nothing
has changed, really. See current stats below.

# cat /proc/interrupts 
           CPU0       CPU1       
  0:   64713875        355    IO-APIC-edge  timer
  2:          0          0          XT-PIC  cascade
  8:          3          1    IO-APIC-edge  rtc
 14:          7          6    IO-APIC-edge  ide0
 16:  176847225  191855106   IO-APIC-level  eth0
 18:     499139     336893   IO-APIC-level  libata
 48:   31491551   22761438   IO-APIC-level  3w-9xxx
NMI:          0          0 
LOC:   64049632   64155206 
ERR:          0
MIS:          0
# uptime 
 08:39:14 up 2 days, 23:54,  8 users,  load average:
0.16, 0.13, 0.07

I have also tried playing with the parameters our friend Ville has mentioned in his post, nothing has come out of it.

I'm willing to give any developer here access to my production machine so that they can see the situation
first hand. Performance is just aweful.

I'm planning to ditch RAID5 on this card and try JBOD and in spreading files evenly across my 12 disks, hopefully this would give some benifit.

Something is very wrong with this card / driver / firmware and or kernel combination, hopefully someone can help out.

Much appreciated

Quoted 29 lines · --- Ian Morgan

Why are you booting with 'noapic'.. in my experience that will seriously
impact interrupt performance. Use the APIC if you've got it, which in this
case you definitely do.

Yes, having your gigabit NIC and RAID controller on the same IRQ (in PIC
mode) could definitely me a source of trouble.

In your web server testing, were you using an external traffic generator or
an on-host process? If you try on-host (eliminating the network throughput
and related interrupts) does performance improve?

So two biggest suggestions:

- Use the APIC. It is your friend.

- It looks like the 3ware card and gigabit nic are on different busses, but
the pirq lines are being routed to the same legacy interrupt in PIC mode. So
APIC mode should avoid that problem. If the controller and nic are actually
on the same bus, separate them.

Quoted 5 lines

Regards,
Ian Morgan

-------------------------------------------------------------------

Quoted 7 lines

Ian E. Morgan Vice President & C.O.O.
Webcon, Inc.
imorgan at webcon dot ca PGP: #2DA40D07
www.webcon.ca
* Customized Linux network solutions for your
business *

-------------------------------------------------------------------

Quoted 118 lines · subbie subbie
Quoted 6 lines

Dear list,

After almost two weeks of experimentation, google searches and reading of posts, bug reports and discussions I'm still far from an answer. I'm

hoping

Quoted 4 lines

someone on this list could shed some light on the subject.

I'm using a 3Ware 9500S-12 card and am able to

produce

Quoted 2 lines

up to 400MB/s sustained read from my 12-disk 4.1TB RAID5 SATA array, 128MB cache onboard, ext3

formatted.

Quoted 5 lines

All is well when performing a single read -- it
works nice and fast.

The system is a web server, serving mid-size files (50MB, each, on average). All hell breaks loose

when

Quoted 1 line

doing many concurrent reads, anywhere between 200

to

Quoted 1 line

400 concurrent streams things simply grind to a

halt

Quoted 5 lines

and the system transfers a maximum of 12-14MB/s.

I'm in the process of clearing up the array (this would take some time) and restructuring it to JBOD mode in order to use each disk individually. I

will

Quoted 2 lines

use a filesystem more suitable to streaming large files, such as XFS. But this would take time and

I

Quoted 1 line

would very much appreciate the advice of people in

the

Quoted 1 line

know if this is going to help at all. It's hard

for

Quoted 3 lines

me to make extreme experimentation (deleting, formatting, reformatting) as this is a productio n system with many files that I have no other place

to

Quoted 5 lines

dump until they can be safely removed. Though I'm working on dumping them slowly to other, remote, machines.

I'm running the latest kernel, 2.6.13.2 and the

latest

Quoted 1 line

3Ware driver, taken from the 3ware.com web site

which

Quoted 1 line

upon insmod, updates the card's firmware to the

latest

Quoted 3 lines

version as well.

In my experiments, I've tried using larger

readahead,

Quoted 2 lines

currently at 16k (this helps, higher values do not seem to help much), using the deadline scheduler

for

Quoted 3 lines

this device, booting the system with the 'noapic' option and playing with a bunch of VM tunable parameters which I'm not sure that I should really

be

Quoted 2 lines

touching. At the moment only the readahead modification is used as the other stuff simply

didn't

Quoted 3 lines

help at all.

With the stock kernel shipped with my

distribution,

Quoted 2 lines

2.6.8 and its old 3ware driver things were just as worse but manifested themselves differently.

The

Quoted 1 line

system was visibly (top, vmstat...) spending most

of

Quoted 10 lines

its time in io-wait and load average was extremely
high, in the area of 10 to 20. With the recent
kernel and driver mentioned above, the excessive io-wait and load seems to have been resolved and observed loadavg is between 1 and 4.

I don't have much experience with systems that are supposed to stream many files concurrently off a hardware RAID of this configuration, but my gut feeling is that something is very wrong and I

should

Quoted 7 lines

be seeing a much higher read throughput.

Trying to preempt people's questions I've tried to include as much information as possible, a lot of stuff is pasted below.

I've just seen that the 3ware driver shares the

same

Quoted 1 line

IRQ with my ethernet card, which has got me a

little

Quoted 10 lines

worried, should I be?

System uptime, exactly 1 day:

# cat /proc/interrupts
          CPU0       CPU1
 0:   21619638          0          XT-PIC  timer
 2:          0          0          XT-PIC  cascade
 8:          4          0          XT-PIC  rtc
10:  268753224          0          XT-PIC 

3w-9xxx,

Quoted 8 lines

eth0

14:         11          0          XT-PIC  ide0
15:     337881          0          XT-PIC  libata
NMI:          0          0
LOC:   21110402   21557685
ERR:          0
MIS:          0
Quoted 43 lines
Quoted 4 lines
# free
           total       used       free     shared
buffers     cached
Mem:       2075260    2024724      50536         

0

Quoted 37 lines

5184 1388408
-/+ buffers/cache: 631132 1444128
Swap: 3903784 0 3903784

# vmstat -n 1  (output of the last few seconds):
procs -----------memory---------- ---swap--
-----io---- --system-- ----cpu----
0  0      0  49932   4760 1392980    0    0 15636
32 3169  3697  4  6 30 60
0  0      0  50816   4752 1392376    0    0  5844
0 3114  3929  3  5 91  1
0  0      0  50924   4772 1391404    0    0  9360
0 3187  4348  6  6 76 13
0  2      0  50552   4780 1391532    0    0 24976
44 4077  3906  3  7 65 25
0  1      0  50444   4780 1392688    0    0 20192
0 5048  3914  7  8 56 30
0  1      0  50568   4756 1392508    0    0 21248
0 4060  3603  4  6 48 41
0  0      0  50704   4724 1392268    0    0 30004
0 3834  3369  4  9 65 22
0  3      0  50556   4728 1392468    0    0  3248
1832 2974  4514  2  5 58 35
0  3      0  50308   4724 1392200    0    0  1288
336 1766  1886  1  3 50 47
0  4      0  50308   4732 1391852    0    0  2300
408 1919  2158  0  3 51 46
0  4      0  50556   4736 1390692    0    0  1856
532 1488  1846  3  1 50 46
0  3      0  50680   4740 1390620    0    0  4016
1296 1577  1682  2  2 50 47
0  3      0  50432   4752 1391628    0    0  2180
72 1730  1945  2  2 51 46
2  2      0  49924   4772 1391540    0    0 44372
564 3403  2847  4  5 50 42
0  0      0  50684   4784 1391528    0    0 28640
216 3804  3847  7  8 69 16

=== message truncated ===

__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com

#3 lore raw
Quoted 3 lines

Something is very wrong with this card / driver / firmware and or kernel combination, hopefully someone can help out.

I think I have to agree; see my post from:

http://marc.theaimsgroup.com/?l=linux-kernel&m=112282837926689&w=2

I've got about 30MB/s from a single threaded version of my backup code - which seems rather on the low side for a modern RAID-5; with multiple writers I was seeing sub-5MB/s but that might be fair if it is seeking everywhere.

I'd be interested to hear how your experiments with jbod'ing them go.

Dave
--
-----Open up your eyes, open up your mind, open up your code -------

/ Dr. David Alan Gilbert    | Running GNU/Linux on Alpha,68K| Happy  \ 
\ gro.gilbert @ treblig.org | MIPS,x86,ARM,SPARC,PPC & HPPA | In Hex /
 \ _________________________|_____ http://www.treblig.org   |_______/
#4 lore raw

A single thread writing at 30MB/s is still not on par with 3ware's specs.

I see that you're also running RAID5 and in this case 3ware did report bad write performance on RAID5 and that was fixed with recent firmwares.

The latest linux driver off their website also includes the latest firmware inside it and flashes the card upon load, make sure to use that.

I'm getting a little over 50MB/s when writing to my RAID volume when completely idle, there's no reason why you should get less.

As for read performance, nothing helps with many concurrent reads, what I get is simply aweful performance no matter what I do.

I'll let you guys know once I try JBOD (as soon as all the data is moved away).

According to Ville answering me privately:

Quoted 1 line

Unfortunately, it's not limited to just that

firmware version or kernel version or driver version. I've tried several firmwares, 2.4.x and 2.6.x kernels and driver version - no salvage.

I do agree, which leads me to believe this is something very specific with the RAID controller itself or its firmware.

Maybe something is so badly designed in this controller that it can't physically do better than that ? Anyone has experience with this controller and its performance on windoze?

Can someone in the know give us some input ?

Thanks

Quoted 8 lines · --- "Dr. David Alan Gilbert
Quoted 2 lines

Something is very wrong with this card / driver /
firmware and or kernel combination, hopefully

someone

Quoted 1 line

can help out.

I think I have to agree; see my post from:

http://marc.theaimsgroup.com/?l=linux-kernel&m=112282837926689&w=2

Quoted 12 lines

I've got about 30MB/s from a single threaded version of my
backup code - which seems rather on the low side for a modern RAID-5; with multiple writers I was seeing sub-5MB/s
but that might be fair if it is seeking everywhere.

I'd be interested to hear how your experiments with jbod'ing them
go.

Quoted 11 lines

Dave
--
-----Open up your eyes, open up your mind, open up
your code -------
/ Dr. David Alan Gilbert | Running GNU/Linux on
Alpha,68K| Happy \
\ gro.gilbert @ treblig.org | MIPS,x86,ARM,SPARC,PPC & HPPA | In Hex /
\ _________________________|_____
http://www.treblig.org |_______/

__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com

#5 lore raw
Quoted 10 lines · * subbie subbie ([email protected]) wrote:

A single thread writing at 30MB/s is still not on par with 3ware's specs.

I see that you're also running RAID5 and in this case 3ware did report bad write performance on RAID5 and that was fixed with recent firmwares.

The latest linux driver off their website also includes the latest firmware inside it and flashes the card upon load, make sure to use that.

I've got driver/firmware that is about 2months old that certainly helped; prior to that I was getting card timeouts (although I also upgraded the e1000 driver at the same time so it might have been that rather than the 3ware that helped).
(Note: I don't expect a driver to perform a dangerous operation like firmware flashing on boot!)

Quoted 3 lines

I'm getting a little over 50MB/s when writing to my RAID volume when completely idle, there's no reason why you should get less.

Well my ~30MB/s is sucking over gig ether and writing in 10MB chunks; but still 50MB/s for RAID5 feels like it sucks.

Quoted 2 lines

I'll let you guys know once I try JBOD (as soon as all the data is moved away).

Nod.

Dave
--
-----Open up your eyes, open up your mind, open up your code -------

/ Dr. David Alan Gilbert    | Running GNU/Linux on Alpha,68K| Happy  \ 
\ gro.gilbert @ treblig.org | MIPS,x86,ARM,SPARC,PPC & HPPA | In Hex /
 \ _________________________|_____ http://www.treblig.org   |_______/
#6 lore raw

Hi

Quoted 7 lines

I've got driver/firmware that is about 2months old that
certainly helped; prior to that I was getting card timeouts (although I also upgraded the e1000 driver at the same time so it might have been that rather than the 3ware that helped).

Two months old might be too old, make sure you have at least 9.2.1.1

Quoted 3 lines

(Note: I don't expect a driver to perform a dangerous
operation like firmware flashing on boot!)

I believe they are writing to NVRAM or similar so that makes it much less risky, otherwise they wouldn't have to write it on each load ... or I might be wrong and they do it once, not sure.

Quoted 11 lines
Quoted 1 line

I'm getting a little over 50MB/s when writing to

my

Quoted 1 line

RAID volume when completely idle, there's no

reason

Quoted 1 line

why you should get less.

Well my ~30MB/s is sucking over gig ether and writing
in 10MB chunks; but still 50MB/s for RAID5 feels like
it sucks.

Yes, that's terrible.

Reading the release notes for the latest one, 9.3.0 that also supports their very latest controller, 9550SX, I get the feeling that write performance is something of an on-going issue with this family of controllers even though they disguise it in this latest driver version as a tweak in write performance.

Another thing, the BBU (battery backup unit) that you are using was also a can of worms for them so this makes it even more important for you to upgrade to the very latest version ,they seem to be having ongoing issues with that one.

Good luck to us all.

__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com

#7 lore raw
In reply to subbie subbie

I have seen some systems on which IRQ load balancing can have a detrimental effect on some devices such as gigabit Ethernet etc.

You could try disabling both the irqbalance userspace daemon (if that's part of your distribution), and in-kernel IRQ balancing, if enabled (CONFIG_IRQBALANCE).

For your NIC, try enabling NAPI interrupt mitigation, if available. This will significantly reduce the interrupt load under high traffic volume.

I guess there's another obvious question that I forgot: Do you have the 3ware cache enabled or disabled? Are your ext3 filesystems mounted with the 'noatime' option?

Regards,
Ian Morgan

Signature · 5 lines
-------------------------------------------------------------------
 Ian E. Morgan          Vice President & C.O.O.       Webcon, Inc.
 imorgan at webcon dot ca        PGP: #2DA40D07      www.webcon.ca
    *  Customized Linux network solutions for your business  *
-------------------------------------------------------------------
Quoted 251 lines · subbie subbie

Dear Ian,

'noapic' was a recommendation by 3Ware / AMCC tech
support. It did not help at all, as expected.
Unfortunately they did not have any other recommendations.

I've now removed 'noapic' and unfortunately nothing
has changed, really. See current stats below.

# cat /proc/interrupts
          CPU0       CPU1
 0:   64713875        355    IO-APIC-edge  timer
 2:          0          0          XT-PIC  cascade
 8:          3          1    IO-APIC-edge  rtc
14:          7          6    IO-APIC-edge  ide0
16:  176847225  191855106   IO-APIC-level  eth0
18:     499139     336893   IO-APIC-level  libata
48:   31491551   22761438   IO-APIC-level  3w-9xxx
NMI:          0          0
LOC:   64049632   64155206
ERR:          0
MIS:          0
# uptime
08:39:14 up 2 days, 23:54,  8 users,  load average:
0.16, 0.13, 0.07

I have also tried playing with the parameters our friend Ville has mentioned in his post, nothing has come out of it.

I'm willing to give any developer here access to my production machine so that they can see the situation
first hand. Performance is just aweful.

I'm planning to ditch RAID5 on this card and try JBOD and in spreading files evenly across my 12 disks, hopefully this would give some benifit.

Something is very wrong with this card / driver / firmware and or kernel combination, hopefully someone can help out.

Much appreciated

Quoted 29 lines · --- Ian Morgan

Why are you booting with 'noapic'.. in my experience that will seriously
impact interrupt performance. Use the APIC if you've got it, which in this
case you definitely do.

Yes, having your gigabit NIC and RAID controller on the same IRQ (in PIC
mode) could definitely me a source of trouble.

In your web server testing, were you using an external traffic generator or
an on-host process? If you try on-host (eliminating the network throughput
and related interrupts) does performance improve?

So two biggest suggestions:

- Use the APIC. It is your friend.

- It looks like the 3ware card and gigabit nic are on different busses, but
the pirq lines are being routed to the same legacy interrupt in PIC mode. So
APIC mode should avoid that problem. If the controller and nic are actually
on the same bus, separate them.

Quoted 5 lines

Regards,
Ian Morgan

--

-------------------------------------------------------------------

Quoted 7 lines

Ian E. Morgan Vice President & C.O.O.
Webcon, Inc.
imorgan at webcon dot ca PGP: #2DA40D07
www.webcon.ca
* Customized Linux network solutions for your
business *

-------------------------------------------------------------------

Quoted 160 lines · subbie subbie
Quoted 6 lines

Dear list,

After almost two weeks of experimentation, google searches and reading of posts, bug reports and discussions I'm still far from an answer. I'm

hoping

Quoted 4 lines

someone on this list could shed some light on the subject.

I'm using a 3Ware 9500S-12 card and am able to

produce

Quoted 2 lines

up to 400MB/s sustained read from my 12-disk 4.1TB RAID5 SATA array, 128MB cache onboard, ext3

formatted.

Quoted 5 lines

All is well when performing a single read -- it
works nice and fast.

The system is a web server, serving mid-size files (50MB, each, on average). All hell breaks loose

when

Quoted 1 line

doing many concurrent reads, anywhere between 200

to

Quoted 1 line

400 concurrent streams things simply grind to a

halt

Quoted 5 lines

and the system transfers a maximum of 12-14MB/s.

I'm in the process of clearing up the array (this would take some time) and restructuring it to JBOD mode in order to use each disk individually. I

will

Quoted 2 lines

use a filesystem more suitable to streaming large files, such as XFS. But this would take time and

I

Quoted 1 line

would very much appreciate the advice of people in

the

Quoted 1 line

know if this is going to help at all. It's hard

for

Quoted 3 lines

me to make extreme experimentation (deleting, formatting, reformatting) as this is a productio n system with many files that I have no other place

to

Quoted 5 lines

dump until they can be safely removed. Though I'm working on dumping them slowly to other, remote, machines.

I'm running the latest kernel, 2.6.13.2 and the

latest

Quoted 1 line

3Ware driver, taken from the 3ware.com web site

which

Quoted 1 line

upon insmod, updates the card's firmware to the

latest

Quoted 3 lines

version as well.

In my experiments, I've tried using larger

readahead,

Quoted 2 lines

currently at 16k (this helps, higher values do not seem to help much), using the deadline scheduler

for

Quoted 3 lines

this device, booting the system with the 'noapic' option and playing with a bunch of VM tunable parameters which I'm not sure that I should really

be

Quoted 2 lines

touching. At the moment only the readahead modification is used as the other stuff simply

didn't

Quoted 3 lines

help at all.

With the stock kernel shipped with my

distribution,

Quoted 2 lines

2.6.8 and its old 3ware driver things were just as worse but manifested themselves differently.

The

Quoted 1 line

system was visibly (top, vmstat...) spending most

of

Quoted 10 lines

its time in io-wait and load average was extremely
high, in the area of 10 to 20. With the recent
kernel and driver mentioned above, the excessive io-wait and load seems to have been resolved and observed loadavg is between 1 and 4.

I don't have much experience with systems that are supposed to stream many files concurrently off a hardware RAID of this configuration, but my gut feeling is that something is very wrong and I

should

Quoted 7 lines

be seeing a much higher read throughput.

Trying to preempt people's questions I've tried to include as much information as possible, a lot of stuff is pasted below.

I've just seen that the 3ware driver shares the

same

Quoted 1 line

IRQ with my ethernet card, which has got me a

little

Quoted 10 lines

worried, should I be?

System uptime, exactly 1 day:

# cat /proc/interrupts
          CPU0       CPU1
 0:   21619638          0          XT-PIC  timer
 2:          0          0          XT-PIC  cascade
 8:          4          0          XT-PIC  rtc
10:  268753224          0          XT-PIC

3w-9xxx,

Quoted 8 lines

eth0

14:         11          0          XT-PIC  ide0
15:     337881          0          XT-PIC  libata
NMI:          0          0
LOC:   21110402   21557685
ERR:          0
MIS:          0
Quoted 4 lines
# free
           total       used       free     shared
buffers     cached
Mem:       2075260    2024724      50536

0

Quoted 37 lines

5184 1388408
-/+ buffers/cache: 631132 1444128
Swap: 3903784 0 3903784

# vmstat -n 1  (output of the last few seconds):
procs -----------memory---------- ---swap--
-----io---- --system-- ----cpu----
0  0      0  49932   4760 1392980    0    0 15636
32 3169  3697  4  6 30 60
0  0      0  50816   4752 1392376    0    0  5844
0 3114  3929  3  5 91  1
0  0      0  50924   4772 1391404    0    0  9360
0 3187  4348  6  6 76 13
0  2      0  50552   4780 1391532    0    0 24976
44 4077  3906  3  7 65 25
0  1      0  50444   4780 1392688    0    0 20192
0 5048  3914  7  8 56 30
0  1      0  50568   4756 1392508    0    0 21248
0 4060  3603  4  6 48 41
0  0      0  50704   4724 1392268    0    0 30004
0 3834  3369  4  9 65 22
0  3      0  50556   4728 1392468    0    0  3248
1832 2974  4514  2  5 58 35
0  3      0  50308   4724 1392200    0    0  1288
336 1766  1886  1  3 50 47
0  4      0  50308   4732 1391852    0    0  2300
408 1919  2158  0  3 51 46
0  4      0  50556   4736 1390692    0    0  1856
532 1488  1846  3  1 50 46
0  3      0  50680   4740 1390620    0    0  4016
1296 1577  1682  2  2 50 47
0  3      0  50432   4752 1391628    0    0  2180
72 1730  1945  2  2 51 46
2  2      0  49924   4772 1391540    0    0 44372
564 3403  2847  4  5 50 42
0  0      0  50684   4784 1391528    0    0 28640
216 3804  3847  7  8 69 16
#8 lore raw
Quoted 9 lines · --- "Ian E. Morgan

I have seen some systems on which IRQ load balancing can have a detrimental
effect on some devices such as gigabit Ethernet etc.

You could try disabling both the irqbalance userspace daemon (if that's part
of your distribution), and in-kernel IRQ balancing, if enabled
(CONFIG_IRQBALANCE).

I don't have a userspace daemon for that, but I'll try the kernel option.

Quoted 5 lines

For your NIC, try enabling NAPI interrupt mitigation, if available. This
will significantly reduce the interrupt load under high traffic volume.

It's always enabled in my configs.

Quoted 5 lines

I guess there's another obvious question that I
forgot: Do you have the
3ware cache enabled or disabled? Are your ext3
filesystems mounted with the
'noatime' option?

Write caching is enabled. I don't have much activity across thousands of files so noatime is less ciritical, but the RAID volume is still mounted noatime.

So basically I'll try the irq load balancing and see whath happens.

Thanks

__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com

#9 lore raw
In reply to subbie subbie
Quoted 7 lines · you [subbie subbie]

Dear list,

After almost two weeks of experimentation, google searches and reading of posts, bug reports and discussions I'm still far from an answer. I'm hoping someone on this list could shed some light on the subject.

Hi,

We're having similar problems with 9500S-4LP and two Hitachi 250GB SATA disks.

Currently we are running 2.6.12.5 and its 3w-9xxx driver. We have tried numerous 2.4 (Red Hat and kernel.org) and 2.6 kernels and 3w-9xxx drivers (kernel and 3ware.com).

The results it more or less the same: on 2.4 it corrupts data and is slow, on 2.6 it doesn't corrupt data, but is slow.

Our workload is VMWare GSX server. Multiple readers/writers will grind the performance to halt.

We've tried:

- http://216.239.59.104/search?q=cache:08R-x7y-Kn8J:xn.pinkhamster.net/blog/tech/3ware_9500_notes.html+blockdev+stra+3w-9xxx&hl=en

      http://www.linux-mag.com/index2.php?option=com_content&task=view&id=2033&Itemid=2304&pop=1&page=0

             blockdev -setra 16384 /dev/sda

- different kernel IO schedulers

- parameters such as
blockdev --setra 8192 /dev/vg0/root
echo 300 > /sys/block/sda/queue/iosched/read_expire echo 5 > /sys/block/sda/queue/iosched/writes_starved

None of this seems to make much difference.

Please drop me a note if you ever get any light on this issue...

-- v --

[email protected]

#10 lore raw
In reply to subbie subbie

In article <[email protected]>,

Quoted 11 lines · subbie subbie

I'm using a 3Ware 9500S-12 card and am able to produce up to 400MB/s sustained read from my 12-disk 4.1TB RAID5 SATA array, 128MB cache onboard, ext3 formatted.
All is well when performing a single read -- it
works nice and fast.

The system is a web server, serving mid-size files (50MB, each, on average). All hell breaks loose when doing many concurrent reads, anywhere between 200 to 400 concurrent streams things simply grind to a halt and the system transfers a maximum of 12-14MB/s.

There are a couple of things you should do:

1. Use the CFQ I/O scheduler, and increase nr-requests:
echo cfq > /sys/block/hda/queue/scheduler echo 1024 > /sys/block/hda/queue/nr_requests

2. Make sure that your filesystem knows about the stripe size and number of disks in the array. E.g. for a raid5 array with a stripe size of 64K and 6 disks (effectively 5, because in every stripe-set there is on disk doing parity):

   # ext3 fs, 5 disks, 64K stripe, units in 4K blocks
   mkfs -text3 -E stride=$((64/4))
   # xfs, 5 disks, 64K stripe, units in 512 bytes
   mkfs -txfs -d sunit=$((64*2)) -d swidth=$((5*64*2))

3. Don't use partitions. Partions do not start on a multiple of the (stripe_size * nr_disks), so your I/O will be misaligned and the settings in (2) will have no or an adverse effect. If you must use partitions, either build them manually with sfdisk so that partitions do start on that multiple, or use LVM.

4. Reconsider your stripe size for streaming large files. If you have say 4 disks, and a 64K
stripe size, then a read of a block of 256K will busy all 4 disks. Many simultaneous threads reading blocks of 256K will result in trashings disks as they all want to read from all 4 disks .. so in that case, using a stripesize of 256K will make things better. One read of 256K (in the ideal, aligned case) will just keep one disk busy. 4 reads can happen in parallel without trashing. Esp. in this case, you need the alignment I talked about in (3).

5. Defragment the files.
If the files are written sequentially, they will not be fragmented. But if they were stored by writing to thousands of them appending a few K at a time in round-robin fashion, you need to defragment.. in the case of XFS, run xfs_fsr every so often.

Good luck,

Mike.