CPUs 2 / cores 4 / threads 16
CPU : | family , model 26 and stepping 5 |
model name : | Intel(R) Xeon(R) CPU X5560 @ 2.80GHz |
cpu MHz : | 1600.000 /2801.000 |
cache size : | 8192 KB |
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 syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 sse4_2 popcnt lahf_lm ida dts tpr_shadow vnmi flexpriority ept vpid |
bogomips : | 5599.19 |
Mainboard : | Supermicro X8DAH+-F_ |
Memory : | 96G DDR3 @ 1333 ECC |
Memory benchmark (hdparm) : | 8823.29 MB/sec |
HD benchmark (hdparm udma6) : (Intel X18-M/X25-M/X25-V G2 SSDs) | 188.78 MB/sec |
OS:openSUSE 11.4
openSSL benchmark:
linux:~> openssl version
OpenSSL 1.0.0k 5 Feb 2013
linux:~> openssl engine
(aesni) Intel AES-NI engine (no-aesni) (dynamic) Dynamic engine loading support
linux:~> openssl speed -evp aes-128-cbc rc4 blowfish
The 'numbers' are in 1000s of bytes per second processed.
type | 16 bytes | 64 bytes | 256 bytes | 1024 bytes | 8192 bytes |
rc4 | 333908.14k | 358351.93k | 322019.06k | 324682.66k | 319383.89k |
blowfish cbc | 108741.83k | 113345.84k | 113786.26k | 111634.15k | 113263.30k |
aes-128-cbc | 86397.43k | 95110.10k | 98164.18k | 98861.21k | 100087.61k |
openSSH benchmark:
Create big empty file (1Gb file)dd if=/dev/zero of=bigfile count=0 seek=2000k
ssh benchmark - to localhost.
scp bigfile localhost:/dev/null for i in aes128-cbc blowfish ; do scp -c bigfile localhost:/dev/null ; done
bigfile.nmRm2q76y1 100% 1000MB 62.5MB/s 00:16 bigfile.nmRm2q76y1 100% 1000MB 47.6MB/s 00:21 bigfile.nmRm2q76y1 100% 1000MB 71.4MB/s 00:14
default : | 62.5MB/s |
aes128-cbc : | 47.6MB/s |
blowfish : | 71.4MB/s |
encrypted ramdisk test:
It is evident, thate there is some distortion due to the write cache: the machine has 96GB ram vs 1.5G ramdisk size.unencrypted | encrypted | |||
blocksize | write speed | read speed | encrypt speed | decrypt speed |
512 | 475.249 | 797.113 | 543.139 | 119.374 |
1k | 788.407 | 1220.117 | 702.260 | 118.573 |
2k | 1053.555 | 1314.705 | 1029.413 | 123.860 |
4k | 1290.156 | 1972.262 | 1286.758 | 122.882 |
8k | 1386.975 | 2218.426 | 1407.089 | 120.780 |
16k | 1451.842 | 2299.632 | 1454.362 | 120.326 |
1M | 1478.007 | 1482.638 | 1486.723 | 120.268 |
size: 1500 megabytes, cipher: aes-cbc-essiv:sha256, key: 256b
In the next test I have specified oflag=direct in the write/encrypt phase of the test:
dd if=/dev/zero of=/mnt/cryptramdisk/testfile bs=$i oflag=direct
unencrypted | encrypted | |||
blocksize | write speed | read speed | encrypt speed | decrypt speed |
512 | 73.683 | 855.768 | 6.802 | 119.000 |
1k | 117.380 | 986.258 | 29.959 | 125.969 |
2k | 213.170 | 1432.815 | 44.993 | 127.687 |
4k | 390.188 | 1584.645 | 74.000 | 127.211 |
8k | 721.074 | 2281.958 | 85.902 | 126.657 |
16k | 1188.749 | 1471.006 | 89.185 | 124.285 |
1M | 2412.269 | 2280.647 | 106.671 | 124.164 |
size: 1500 megabytes, cipher: aes-cbc-essiv:sha256, key: 256b, dd: oflag=direct
Žádné komentáře:
Okomentovat