[Michael T. Babcock] XEN Benchmarks


I installed FC6 with XEN on a new server (dual Opteron 2210 dual-core processors) and wanted to test some speed differences with and without XEN running. I knew I'd be putting the machine into production soon, so I thought if I ran a few quick tests I might be able to provide some useful information for others as not much benchmark data was available. Hopefully this is interesting or helpful to someone. Feel free to send me E-mail to mike at this domain if you have any questions or comments (or related benchmarks to link to).

Distributed.net

I ran the quick and dirty dnetc -bench benchmark to test the basic performance of the CPUs on this system both in the xen enabled kernel in Dom0 and in the bare kernel with no xen features.

-dnetc v2.9011-496-CFR-05060815 for Linux (Linux 2.6.20-1.2933.fc6).
+dnetc v2.9011-496-CFR-05060815 for Linux (Linux 2.6.20-1.2933.fc6xen).

  RC5-72: using core #0 (SNJL 3-pipe).
  RC5-72: Benchmark for core #0 (SNJL 3-pipe)
-                      0.00:00:16.60
+                      0.00:00:16.99
  RC5-72: using core #1 (KBE-64 3-pipe).
  RC5-72: Benchmark for core #1 (KBE-64 3-pipe)
-                      0.00:00:16.36
+                      0.00:00:16.10
  RC5-72: using core #2 (ANSI 4-pipe).
  RC5-72: Benchmark for core #2 (ANSI 4-pipe)
-                      0.00:00:16.15
+                      0.00:00:16.62
  RC5-72: using core #3 (ANSI 2-pipe).
  RC5-72: Benchmark for core #3 (ANSI 2-pipe)
-                      0.00:00:16.35
+                      0.00:00:17.52
  RC5-72: using core #4 (ANSI 1-pipe).
  RC5-72: Benchmark for core #4 (ANSI 1-pipe)
-                      0.00:00:16.58
+                      0.00:00:17.36
  OGR-P2: Benchmark for core #0 (GARSP 6.0-64)
-                      0.00:00:16.51
+                      0.00:00:16.42

I'm not sure what exactly the margin of error for these tests is -- I may have to run them a few more times to find out -- but it would appear the xen kernel sometimes is faster and the non-xen kernel is faster the other times.

For the record, I had almost nothing running but the basics were there -- ssh (just me), portmap, hald and ntpd.

Drive speed

I've configured my four SATA drives as a RAID-5 set in Dom0 and ran Bonnie++ on it to test performance. Output from dmesg shows:

ata1: SATA max UDMA/133 cmd 0xFFFFC20000024000 ctl 0xFFFFC20000024020 bmdma 0xFFFFC20000024030 irq 11
ata2: SATA max UDMA/133 cmd 0xFFFFC20000024100 ctl 0xFFFFC20000024120 bmdma 0xFFFFC20000024130 irq 11
ata3: SATA max UDMA/133 cmd 0xFFFFC20000024200 ctl 0xFFFFC20000024220 bmdma 0xFFFFC20000024230 irq 11
ata4: SATA max UDMA/133 cmd 0xFFFFC20000024300 ctl 0xFFFFC20000024320 bmdma 0xFFFFC20000024330 irq 11
scsi0 : sata_svw
ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata1.00: ATA-7, max UDMA/133, 488397168 sectors: LBA48 NCQ (depth 0/32)
ata1.00: ata1: dev 0 multi count 16
ata1.00: configured for UDMA/133
scsi1 : sata_svw
ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata2.00: ATA-7, max UDMA/133, 488397168 sectors: LBA48 NCQ (depth 0/32)
ata2.00: ata2: dev 0 multi count 16
ata2.00: configured for UDMA/133
scsi2 : sata_svw
ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata3.00: ATA-7, max UDMA/133, 488397168 sectors: LBA48 NCQ (depth 0/32)
ata3.00: ata3: dev 0 multi count 16
ata3.00: configured for UDMA/133
scsi3 : sata_svw
ata4: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata4.00: ATA-7, max UDMA/133, 488397168 sectors: LBA48 NCQ (depth 0/32)
ata4.00: ata4: dev 0 multi count 16
ata4.00: configured for UDMA/133

scsi 0:0:0:0: Direct-Access ATA ST3250820NS 3.AE PQ: 0 ANSI: 5 SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB) sda: Write Protect is off sda: Mode Sense: 00 3a 00 00 SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB) sda: Write Protect is off sda: Mode Sense: 00 3a 00 00 SCSI device sda: write cache: enabled, read cache: enabled, doesn't support DPO or FUA sda: sda1 sda2 sda3 sd 0:0:0:0: Attached scsi disk sda
(repeat that last part three more times for each sdb, sdc and sdd)

Bonnie was run as: bonnie++ -d /var/www/html/bonnie -s 5000 -m quadopt_domU -n 1000 -u 99 -x 10 although in both cases below I cut it off after less than 10 iterations to get other work done.

Raw Bonnie++ output is as follows from within Dom0 on an idle machine with no DomUs, under LVM:

name,file_size,putc,putc_cpu,put_block,put_block_cpu,rewrite,rewrite_cpu,getc,getc_cpu,get_block,get_block_cpu,seeks,seeks_cpu,num_files,seq_create,seq_create_cpu,seq_stat,seq_stat_cpu,seq_del,seq_del_cpu,ran_create,ran_create_cpu,ran_stat,ran_stat_cpu,ran_del,ran_del_cpu
quadopt,5000M,50131,96,50260,13,31130,7,44048,85,159690,28,273.3,0,1000,52,99,25708,36,31317,53,51,99,115,81,110,91
quadopt,5000M,47109,91,50728,13,34683,8,43977,85,158069,28,270.6,0,1000,53,99,23922,33,29881,50,52,99,113,79,106,91
quadopt,5000M,46272,90,51006,13,34880,8,43838,85,159635,27,267.5,0,1000,53,99,21742,30,30604,51,48,99,109,76,110,91
quadopt,5000M,46660,90,50861,14,35302,8,42890,85,158718,27,266.9,0,1000,50,99,19205,27,29738,50,51,99,107,73,110,89

Raw Bonnie++ output on DomU with the same parameters (LVM2 LV in DomU as part of PV hosted by Xen which is really LV in Dom0) was:

name,file_size,putc,putc_cpu,put_block,put_block_cpu,rewrite,rewrite_cpu,getc,getc_cpu,get_block,get_block_cpu,seeks,seeks_cpu,num_files,seq_create,seq_create_cpu,seq_stat,seq_stat_cpu,seq_del,seq_del_cpu,ran_create,ran_create_cpu,ran_stat,ran_stat_cpu,ran_del,ran_del_cpu
quadopt_domU,5000M,38290,82,40828,13,29105,2,42697,53,112871,2,194.5,0,1000,8003,38,696,0,601,0,6180,30,532,0,318,0
quadopt_domU,5000M,31075,66,40588,13,29033,2,42810,52,111821,2,194.3,0,1000,10392,51,982,0,748,1,6756,33,710,0,361,0
quadopt_domU,5000M,35591,77,41021,13,29391,2,42723,52,113911,2,193.2,0,1000,11049,53,1051,0,784,1,9690,46,727,0,350,0
quadopt_domU,5000M,30129,65,40697,13,28888,2,41891,53,113253,2,194.2,0,1000,12140,60,1142,0,786,1,10046,48,786,0,370,0
quadopt_domU,5000M,31205,66,40953,14,28986,2,42047,53,112241,2,193.5,0,1000,11590,55,1074,0,766,1,8698,41,780,0,356,0
quadopt_domU,5000M,26187,56,39823,13,28886,2,42494,51,112749,2,185.3,0,1000,11760,57,889,0,606,1,6191,31,769,0,376,0
quadopt_domU,5000M,23953,51,39890,13,29203,2,40771,53,111786,2,193.4,0,1000,11223,56,1069,0,731,1,9129,46,780,0,348,0
quadopt_domU,5000M,30671,65,41270,13,28789,2,39844,57,112930,2,195.3,0,1000,11105,56,1074,0,752,1,9120,44,775,0,365,0

Raw Bonnie++ output on DomU with the same parameters (but now xvdb1 in DomU as part of PV hosted by Xen which is really LV in Dom0) was:

name,file_size,putc,putc_cpu,put_block,put_block_cpu,rewrite,rewrite_cpu,getc,getc_cpu,get_block,get_block_cpu,seeks,seeks_cpu,num_files,seq_create,seq_create_cpu,seq_stat,seq_stat_cpu,seq_del,seq_del_cpu,ran_create,ran_create_cpu,ran_stat,ran_stat_cpu,ran_del,ran_del_cpu
quadopt_domU,5000M,37315,80,34635,10,29132,3,42996,68,120472,3,202.2,0,1000,12512,69,642,0,372,0,7015,38,537,0,190,0
quadopt_domU,5000M,27631,58,40582,12,29383,3,42943,69,121830,2,203.1,0,1000,12434,68,1167,0,389,0,11539,65,709,0,191,0
Google
 
Stumble it! XFN Friendly Powered by DJBDNS Powered by Zope Valid CSS! Website Security Test

Served by:  Zope 2.7.6

Page Copyright © 2014, Michael T. Babcock. All Rights Reserved.

To contact me, send an E-mail to sawyoursite at this domain.

If you'd really like your mail server reported for spam, send me some junk mail to junk-yum@mikebabcock.me or devnull@mikebabcock.me. This site powered by djbdns