Aktuell gibt es bei NetCup wieder Sonderangebote, wie es sie bei NetCup häufiger gibt.
Einerseits kann man zwischen verschiedenen Größenpaketen wählen, als auch zwischen jeweils zwischen einer SSD und einer normalen SAS HDD.
Leider hab ich selbst keine Benchmarks dazu gefunden, beziehungsweise dann nur entweder von unterschiedlichen Servern oder sie waren sich selber nicht ganz sicher ob die Werste so stimmen können.
Also ganz einfach: selber messen. Dazu verwende ich die aktuellen Server (RS 2000) mit und ohne SSD, als auch meinen alten NetCup Server (M SSD v6) der ein Jahr alt ist.
Zuerst mit hdparm, mit und ohne Cache (-t -T), mit und ohne Kernel Page cache (–direct).
SAS – RS 2000 SAS G7SEa1
root@SAS:~# hdparm -tT --direct /dev/sda1
/dev/sda1:
Timing O_DIRECT cached reads: 2676 MB in 2.00 seconds = 1337.91 MB/sec
Timing O_DIRECT disk reads: 8 MB in 0.05 seconds = 148.82 MB/sec
root@SAS:~# hdparm -tT /dev/sda1
/dev/sda1:
Timing cached reads: 13612 MB in 2.00 seconds = 6811.80 MB/sec
Timing buffered disk reads: 8 MB in 0.19 seconds = 41.73 MB/sec
root@SAS:~# hdparm -tT --direct /dev/sda1
/dev/sda1:
Timing O_DIRECT cached reads: 2780 MB in 2.00 seconds = 1390.03 MB/sec
Timing O_DIRECT disk reads: 8 MB in 0.09 seconds = 85.27 MB/sec
root@SAS:~# hdparm -tT /dev/sda1
/dev/sda1:
Timing cached reads: 15900 MB in 2.00 seconds = 7956.80 MB/sec
Timing buffered disk reads: 8 MB in 0.21 seconds = 38.39 MB/sec
SSD – RS 2000 SSD G7SEa1
root@SSD:~# hdparm -tT --direct /dev/sda1
/dev/sda1:
Timing O_DIRECT cached reads: 1974 MB in 2.00 seconds = 986.61 MB/sec
Timing O_DIRECT disk reads: 4450 MB in 3.00 seconds = 1482.72 MB/sec
root@SSD:~# hdparm -tT /dev/sda1
/dev/sda1:
Timing cached reads: 13782 MB in 2.00 seconds = 6898.12 MB/sec
Timing buffered disk reads: 2052 MB in 3.00 seconds = 683.34 MB/sec
old SSD – Root-Server M SSD v6
root@SSD:~# hdparm -tT --direct /dev/vda1
/dev/vda1:
Timing O_DIRECT cached reads: 938 MB in 2.01 seconds = 467.00 MB/sec
Timing O_DIRECT disk reads: 66 MB in 3.13 seconds = 21.07 MB/sec
root@SSD:~# hdparm -tT /dev/vda1
/dev/vda1:
Timing cached reads: 16792 MB in 2.00 seconds = 8404.16 MB/sec
Timing buffered disk reads: 22 MB in 3.52 seconds = 6.25 MB/sec
Anschließend messen wir noch mal etwas genauer nach mit fio (Optionen aus dem Netcup-Thread genommen).
SAS – RS 2000 SAS G7SEa1
root@SAS:~# fio --randrepeat=1 --ioengine=libaio --direct=1 --gtod_reduce=1 --name=test --filename=test --bs=4k --iodepth=64 --size=4G --readwrite=randrw --rwmixread=75
test: (g=0): rw=randrw, bs=4K-4K/4K-4K/4K-4K, ioengine=libaio, iodepth=64
fio-2.1.11
Starting 1 process
Jobs: 1 (f=1): [m(1)] [100.0% done] [5454KB/1726KB/0KB /s] [1363/431/0 iops] [eta 00m:00s]
test: (groupid=0, jobs=1): err= 0: pid=2012: Sat Jan 21 14:18:58 2017
read : io=3071.7MB, bw=6406.4KB/s, iops=1601, runt=490984msec
write: io=1024.4MB, bw=2136.4KB/s, iops=534, runt=490984msec
cpu : usr=1.37%, sys=6.95%, ctx=890167, majf=0, minf=7
IO depths : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=0.1%, 32=0.1%, >=64=100.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.1%, >=64=0.0%
issued : total=r=786347/w=262229/d=0, short=r=0/w=0/d=0
latency : target=0, window=0, percentile=100.00%, depth=64
Run status group 0 (all jobs):
READ: io=3071.7MB, aggrb=6406KB/s, minb=6406KB/s, maxb=6406KB/s, mint=490984msec, maxt=490984msec
WRITE: io=1024.4MB, aggrb=2136KB/s, minb=2136KB/s, maxb=2136KB/s, mint=490984msec, maxt=490984msec
Disk stats (read/write):
sda: ios=786285/262345, merge=0/72, ticks=30897848/1451668, in_queue=34153792, util=100.00%
SSD – RS 2000 SSD G7SEa1
-
old SSD – Root-Server M SSD v6
root@SSD:~# fio --randrepeat=1 --ioengine=libaio --direct=1 --gtod_reduce=1 --name=test --filename=test --bs=4k --iodepth=64 --size=4G --readwrite=randrw --rwmixread=75 !1839
test: (g=0): rw=randrw, bs=4K-4K/4K-4K/4K-4K, ioengine=libaio, iodepth=64
fio-2.1.11
Starting 1 process
test: Laying out IO file(s) (1 file(s) / 4096MB)
Jobs: 1 (f=1): [m(1)] [100.0% done] [1188KB/388KB/0KB /s] [297/97/0 iops] [eta 00m:05s]
test: (groupid=0, jobs=1): err= 0: pid=2742: Sat Jan 21 23:44:40 2017
read : io=3071.7MB, bw=266816B/s, iops=65, runt=12071512msec
write: io=1024.4MB, bw=88977B/s, iops=21, runt=12071512msec
cpu : usr=0.04%, sys=0.09%, ctx=35579, majf=0, minf=6
IO depths : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=0.1%, 32=0.1%, >=64=100.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.1%, >=64=0.0%
issued : total=r=786347/w=262229/d=0, short=r=0/w=0/d=0
latency : target=0, window=0, percentile=100.00%, depth=64
Run status group 0 (all jobs):
READ: io=3071.7MB, aggrb=260KB/s, minb=260KB/s, maxb=260KB/s, mint=12071512msec, maxt=12071512msec
WRITE: io=1024.4MB, aggrb=86KB/s, minb=86KB/s, maxb=86KB/s, mint=12071512msec, maxt=12071512msec
Disk stats (read/write):
vda: ios=786522/269049, merge=0/6, ticks=579569992/187883904, in_queue=767498600, util=100.00%
Vielen Dank an den Besitzer des SSD Servers für die Benchmarks.