Tanya tentang IO speed


Status
Not open for further replies.
merkurius

merkurius

Poster 2.0
ouh gitu yah..kemarin udah sempet chat2 dengan providernya dan memank sepertiny ada masalah dengan vps yg kvm soal io speed akhirny udah beres juga.... saya udah dapat speed yg normal :D
 
perdhanahost

perdhanahost

Hosting Guru
Ikutan iseng ahh ...

Disk : 4 x 1TB WDC Caviar Black 64MB 7200RPM
RAID : Software (Linux mdadm) RAID10
Sistem : Supermicro

# dd if=/dev/zero of=test.file bs=64k count=16k conv=fdatasync;rm -f test.file
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 13.7926 seconds, 77.8 MB/s
Click to expand...

Disk : 4 x 1TB WDC Caviar Black 64MB 7200RPM
RAID : Software (Linux mdadm) RAID10
Sistem : Intel Server

# dd if=/dev/zero of=test.file bs=64k count=16k conv=fdatasync;rm -f test.file
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 9.34742 seconds, 115 MB/s
Click to expand...

Disk : 2 x 1TB Seagate Constellation.2 ST91000640NS 64MB 7200RPM
RAID : Software (Linux mdadm) RAID1
Sistem : Supermicro Server

# dd if=/dev/zero of=test.file bs=64k count=16k conv=fdatasync;rm -f test.file
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 14.5631 seconds, 73.7 MB/s
Click to expand...

Note: Ketiganya dalam kondisi normal run (tidak kosong dan tidak full load/peak - 70%-80% utilized)

Semoga berguna
 
cpserv

cpserv

Expert 1.0
pakde perdhanahost, coba check /sys/block/XdY/queue/scheduler (dimana X = type hard disknya dan Y = "nomer" hard disknya), kayaqnya masih bisa di tweak lagi tuh software raidnya.. :D coba aja ganti cfq ke deadline atau noop di node, dan sebaliknya di vps juga..
 
deRegen

deRegen

Apprentice 1.0
cpserv said:
pakde perdhanahost, coba check /sys/block/XdY/queue/scheduler (dimana X = type hard disknya dan Y = "nomer" hard disknya), kayaqnya masih bisa di tweak lagi tuh software raidnya.. :D coba aja ganti cfq ke deadline atau noop di node, dan sebaliknya di vps juga..
Click to expand...

SanDisk Ultra SSD, 60GB, SATA II

$ time sh -c "dd if=/dev/zero of=iotest bs=8k count=500000 && sync"
500000+0 records in
500000+0 records out
4096000000 bytes (4.1 GB) copied, 16.993 s, 241 MB/s

utilisasi server masih <50%

VPS berada di DS3200 IBM dengan SAS Cable 3Gps, SATA hard disk. Utilisasi hampir 70%

$ time sh -c "dd if=/dev/zero of=iotest bs=8k count=100000 && sync"

100000+0 records in
100000+0 records out
819200000 bytes (819 MB) copied, 38.219 s, 21.4 MB/s

real 0m43.270s
user 0m0.024s
sys 0m0.816s

$ cat /sys/block/sda/queue/scheduler
noop anticipatory deadline [cfq]

hmm...
 
idroot

idroot

Apprentice 2.0
nyoba test juga disalah satu vps

2x1 tb - raid1 software

root@template:~# dd if=/dev/zero of=iotest bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 10.6341 s, 101 MB/s
 
O

optimus

Beginner 1.0
ikutan nimbrung,

SAS 15000 rpm, pake esxi

[root@rhel6 tmp]# dd if=/dev/zero of=test.file bs=64k count=16k conv=fdatasync;rm -f test.file
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 4.15903 s, 258 MB/s
[root@rhel6 tmp]# time sh -c "dd if=/dev/zero of=iotest bs=8k count=500000 && sync"
500000+0 records in
500000+0 records out
4096000000 bytes (4.1 GB) copied, 12.4123 s, 330 MB/s

sekrinsut

imageshack.us/photo/my-images/443/mbuh.jpg/
 
Status
Not open for further replies.

Top