[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Ultra 5 disk performance [Was: More storage for my Ultra10]



Hi,

having red Andrews mail I tried some disk IO on my Sun Ultra 5
(CPU: 333MHz, RAM: 512MB, kernel: 2.6.12.4)

The system was equipped with its original IDE disk and a newer 200GB
one (btw: all 200gigs are available with newer 2.6 kernels).

Additionally the system is equipped as follows

2) SCSI-system:
SymBios 53C876 controller, (~3 yrs. old) external 30GB SCSI disk (ext3 FS)

3) USB-System:
NEC Corporation USB 2.0 controller, Trekstor USB 2.0 160 GB disk (vfat FS)

4) Silicon Image, Serial ATA (Sil 3112) Controller, Maxtor 200G disk (ext3 FS)

This is what I get when running hdparm -tT /...devpath
(I'll include the result of a bonnie++ at the end of the mail.)

1) IDE
/dev/hda1:
Timing O_DIRECT cached reads:   700 MB in  2.00 seconds = 349.70 MB/sec
Timing O_DIRECT disk reads:   38 MB in  3.09 seconds =  12.31 MB/sec
2) SCSI
/dev/sdb1:
Timing O_DIRECT cached reads:   672 MB in  2.01 seconds = 335.05 MB/sec
Timing O_DIRECT disk reads:   42 MB in  3.07 seconds =  13.66 MB/sec
3) USB
/dev/sdd1:
Timing O_DIRECT cached reads:   660 MB in  2.01 seconds = 328.74 MB/sec
Timing O_DIRECT disk reads:   28 MB in  3.03 seconds =   9.23 MB/sec
4) SATA
/dev/sdc1:
Timing O_DIRECT cached reads:   704 MB in  2.01 seconds = 350.48 MB/sec
HDIO_DRIVE_CMD(null) (wait for flush complete) failed: Inappropriate ioctl for d
evice
Timing O_DIRECT disk reads:  156 MB in  3.01 seconds =  51.82 MB/sec
HDIO_DRIVE_CMD(null) (wait for flush complete) failed: Inappropriate ioctl for d
evice

Regardless of the 'failed' messages for SATA, the disc seems to work fine for me.
I'm not shure about their  relevance  ...

It's not a fair comparison due to older IDE and SCSI hardware, different filesystem on USB et. cetera. Nevertheless it may give You a hint. As I am the one who was really unsuccessful with the selected USB cards I'd see SATA or SCSI as choice for buying new disks.

BTW: I'd be really interested if someone has newer SCSI hardware available for testing.
If money doesn't matter ... :-)

Cheers, Roger

========== Results of bonnie++ (1.03) follows ========
=======================================
# hdparm -tT /dev/hda1 (IDE)

/dev/hda1:
Timing O_DIRECT cached reads:   700 MB in  2.00 seconds = 349.70 MB/sec
Timing O_DIRECT disk reads:   38 MB in  3.09 seconds =  12.31 MB/sec

Bonnie++:
Version 1.03 ------Sequential Output------ --Sequential Input- --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP sunny 1G 2850 99 14461 15 5672 12 2837 96 11758 11 93.6 1 ------Sequential Create------ --------Random Create-------- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 557 99 +++++ +++ 22621 99 555 99 +++++ +++ 2065 99
sunny,1G,2850,99,14461,15,5672,12,2837,96,11758,11,93.6,1,16,557,99,+++++,+++,22621,99,555,99,+++++,+++,2065,99

=======================================
# hdparm -tT /dev/sdb1 (SCSI)

/dev/sdb1:
Timing O_DIRECT cached reads:   672 MB in  2.01 seconds = 335.05 MB/sec
Timing O_DIRECT disk reads:   42 MB in  3.07 seconds =  13.66 MB/sec

Bonnie++:
Version 1.03 ------Sequential Output------ --Sequential Input- --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP sunny 1G 2891 99 20214 20 7715 14 2921 99 14832 12 129.7 2 ------Sequential Create------ --------Random Create-------- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 485 99 +++++ +++ 23127 99 577 99 +++++ +++ 2116 99
sunny,1G,2891,99,20214,20,7715,14,2921,99,14832,12,129.7,2,16,485,99,+++++,+++,23127,99,577,99,+++++,+++,2116,99

=======================================
# hdparm -tT /dev/sdd1 (USB 2.0)

/dev/sdd1:
Timing O_DIRECT cached reads:   660 MB in  2.01 seconds = 328.74 MB/sec
Timing O_DIRECT disk reads:   28 MB in  3.03 seconds =   9.23 MB/sec

Bonnie++:
Version 1.03 ------Sequential Output------ --Sequential Input- --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP sunny 1G 2575 94 11984 39 5247 25 2610 96 11729 30 105.0 9 ------Sequential Create------ --------Random Create-------- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 6 99 14 99 86 99 10 98 14 99 24 99
sunny,1G,2575,94,11984,39,5247,25,2610,96,11729,30,105.0,9,16,6,99,14,99,86,99,10,98,14,99,24,99

=======================================
# hdparm -tT /dev/sdc1 (S-ATA)

/dev/sdc1:
Timing O_DIRECT cached reads:   704 MB in  2.01 seconds = 350.48 MB/sec
HDIO_DRIVE_CMD(null) (wait for flush complete) failed: Inappropriate ioctl for device
Timing O_DIRECT disk reads:  156 MB in  3.01 seconds =  51.82 MB/sec
HDIO_DRIVE_CMD(null) (wait for flush complete) failed: Inappropriate ioctl for device

Version 1.03 ------Sequential Output------ --Sequential Input- --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP sunny 1G 2796 98 39997 83 21047 53 2912 99 37068 34 153.1 2 ------Sequential Create------ --------Random Create-------- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 312 99 +++++ +++ 11976 96 319 99 +++++ +++ 1362 98
sunny,1G,2796,98,39997,83,21047,53,2912,99,37068,34,153.1,2,16,312,99,+++++,+++,11976,96,319,99,+++++,+++,1362,98



begin:vcard
fn:Dr. Roger Zimmermann
n:Zimmermann;Dr. Roger
org;quoted-printable;quoted-printable:Zentrum f=C3=BCr exp. Medizin;Institut f=C3=BCr Neuro- und Pathophysiologie
adr;quoted-printable:;;Martinistra=C3=9Fe 52;Hamburg;;20246;Germany
email;internet:R.Zimmermann@UKE.Uni-Hamburg.de
tel;work:..49 (0)40 42803 5351
tel;fax:..49 (0)40 42803 7255
x-mozilla-html:FALSE
url:http://www.uke.uni-hamburg.de
version:2.1
end:vcard


Reply to: