[4007] in linux-scsi channel archive
raid0 performance
daemon@ATHENA.MIT.EDU (Ulisses Alonso)
Tue May 26 04:58:35 1998
Date: Tue, 26 May 1998 08:54:41 +0000 (GMT)
From: Ulisses Alonso <ulisses@pusa.eleinf.uv.es>
Reply-To: ulisses@pusa.eleinf.uv.es
To: linux-scsi@vger.rutgers.edu
cc: linux-raid@vger.rutgers.edu
Hello,
(Sorry to people in linux-raid@ if the previous email arrived, I'm not
sure on that)
First of all I would like to thank the great job that people is doing
in the linux raid modules
I have made some tests, I'm not sure if are good measurements or
significative. Also I'm not an experienced SCSI user so I can not say if
the I/O system is working fast or slow, at least is faster enough.
I would like to hear some comments about the results...
The system:
Acer AcerAltos 9100
One PII 300Mhz
256Mb
Two onboard Adaptec 2940UW - aic7880
Two IBM 7200rpm (Ultrastar 9ES, DDRS-39130W) on the same controller
(I will use the other channel to plug/unplug DAT/CDR devices while
system is up)
These disks seems to be ultra wide(http://www.storage.ibm.com/hardsoft/diskdrdl/ultra/9esprod.htm),
"not just" wide as the model name
$ cat /proc/scsi/aic7xxx/0
Adaptec AIC7xxx driver version: 4.1/3.2
Compile Options:
AIC7XXX_RESET_DELAY : 15
AIC7XXX_TAGGED_QUEUEING: Enabled
AIC7XXX_PAGE_ENABLE : Enabled
AIC7XXX_PROC_STATS : Disabled
Adapter Configuration:
SCSI Adapter: Adaptec AIC-7880 Ultra SCSI host adapter
(AIC-788x chipset)
Host Bus: Wide
Base IO: 0x9400
Base IO Memory: 0x13100000
IRQ: 15
SCBs: Used 16, HW 16, Page 255
Interrupts: 645456
Serial EEPROM: False
Extended Translation: Disabled
SCSI Bus Reset: Enabled
Ultra SCSI: Enabled <----
Disconnect Enable Flags: 0xffff
The filesystem configuration:
$ mount
/dev/sda1 on / type ext2 (rw,errors=remount-ro)
proc on /proc type proc (rw)
/dev/md0 on /var type ext2 (rw,errors=remount-ro)
/dev/md1 on /home type ext2 (rw,errors=remount-ro)
/dev/md2 on /var/www type ext2 (rw,errors=remount-ro)
The tests:
First I created a file of 10240000 bytes, with this file I
- Copy from non-raid to non-raid fs on the same disk
- Copy from non-raid to raid fs
- Copy from raid fs to raid fs (same disks)
I repeated each test several times, I annotate the best results. The
system was always using the same amount of memory for caching & buffering:
root@caligula:/home/ulisses# free
total used free shared buffers cached
Mem: 257364 253624 3740 57292 62224 113924
-/+ buffers/cache: 77476 179888
Swap: 261496 160 261336
The results:
Non raid to non raid in the same disk
--------------------------------------
root@caligula:/home/ulisses# time cp /tmp/file /tmp/copy
real 0m16.696s
user 0m0.020s
sys 0m4.130s
Non raid to raid on the same disks
----------------------------------
root@caligula:/home/ulisses# time cp /tmp/file /var
real 0m7.903s
user 0m0.080s
sys 0m4.190s
raid to raid on the same disks
------------------------------
root@caligula:/home/ulisses# time cp /var/file /var/copy
real 0m11.253s
user 0m0.030s
sys 0m4.500s
Also hdparm (best)results:
root@caligula:/home/ulisses# hdparm -t /dev/md0
/dev/md0:
Timing buffered disk reads: 32 MB in 2.90 seconds =11.03 MB/sec
root@caligula:/home/ulisses# hdparm -T /dev/md0
/dev/md0:
Timing buffer-cache reads: 64 MB in 1.32 seconds =48.48 MB/sec
root@caligula:/home/ulisses# hdparm -t /dev/sda
/dev/sda:
Timing buffered disk reads: 32 MB in 2.50 seconds =12.80 MB/sec
root@caligula:/home/ulisses# hdparm -T /dev/sda
/dev/sda:
Timing buffer-cache reads: 64 MB in 1.34 seconds =47.76 MB/sec
hdparm test results doesn't differ from test to test...
cp'ing files yes: +-2seg
Any comment will be greatly appreciated
regards,
Ulisses
-----------------------------------------------------------------------------
"Computers are useless. They can only give answers." Pablo Picasso
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
