Performance Summary:

Power Summary:

Power Details:

Validation Errors:

•   No Validation Errors Found

Power Measurement Configuration:

Physical Configuration:

Virtual Configuration:

Notes:

Physical System Notes

• Firmware settings used:
• Memory.PagePolicy=Adaptive
• OperatingModes.ChooseOperatingMode=Maximum Performance
• Power.PlatformControlledType=Maximum Performance
• Processors.C1EnhancedMode=Disable
• Processors.CPUPstateControl=None
• Processors.EnergyEfficientTurbo=Disable
• Processors.L0p=Disable
• Processors.L1=Disable
• Processors.MONITORMWAIT=Enable
• Disable adapter option ROM on PCI slots with network adapters installed.

Storage Notes

• The System Under Test has 24 800GB internal SSDs configured as two RAID5 arrays with twelve disks each.
• One of the arrays contains the logical volume for the hypervisor OS that is booted.
• Each RAID disk has a large parition that is made into a volume group. The disks for the VMs are each allocated as a logical volume from one of the volume groups.

Virtualization Software Notes

• The following kernel boot options were used:
• audit=0 elevator=deadline hpet=disable clocksource=tsc tsc=reliable nohz=on highres=off cgroup_disable=cpu,cpuacct,memory,blkio,hugetlb intel_iommu=on iommu=pt pci=realloc nosoftlockup intel_idle.max_cstate=0 mce=ignore_ce nmi_watchdog=0 ipv6.disable=1 default_hugepagesz=1G hugepagesz=1G
• The following services were disabled after OS installation:
• irqbalance, postfix, smartd
• The following items were added in /usr/lib/sysctl.d/99-specvirt.conf:
• fs.aio-max-nr = 3145728
• kernel.perf_event_max_sample_rate = 1000
• kernel.sched_compat_yield = 1
• kernel.sched_min_granularity_ns = 10000000
• kernel.sched_wakeup_granularity_ns = 15000000
• kernel.sched_latency_ns = 20000000
• kernel.sched_migration_cost_ns = 150000
• kernel.numa_balancing = 0
• vm.dirty_background_ratio = 5
• vm.dirty_ratio = 10
• vm.swappiness = 0
• Configuration file /etc/modprobe.d/i40e.conf includes:
• blacklist i40evf
• options i40e max_vfs=0,0,64,64,64,64,64,64,64,64,64,64,64,64,64,64,64,64
• Network configuration is as follows:
• Ethernet devices eth2, eth3, eth4, eth5, eth6, eth7, eth8, eth9, eth10, eth11, eth12, eth13, eth14, and eth15 are the physical functions that are used by the host to provide physical transport for the virtual functions.
• Kernel module i40e does not have a max_vfs option and ignores it. Instead, a systemd initialization script reads the max_vfs option from /etc/modprobe.d/i40e.conf and allocates the virtual functions.
• 609 virtual functions were used for all the VMs.
• App, Db, Web, and Batch VMs are assigned a VF that is routed to the NUMA node to which the VM's VCPUs are pinned. Infra and Mail VMs are assigned VFs round-robin from the physical functions listed above.
• After the system booted, 1GB huge pages were manually allocated on the NUMA nodes. 336 hugepages were each allocated on nodes 0 and 2. 358 huge pages were allocated on node 1. 337 huge pages were each allocated on nodes 4, 5, 6, and 7.
• VM configurations are as follows:
• All VMs are managed with libvirt and created using the virsh utility.
• All VMs are created with the following XML template:
  •   <domain type='kvm'>
  •   <name>$vm_name</name>
  •   <memory unit='MiB'>$vm_memory</memory>
  •   <vcpu cpuset='$host-cpu-list'>$vm_vcpus</vcpu>
  •   <os>
  •     <type arch='x86_64' machine='pc'>hvm</type>
  •     <boot dev='hd'/>
  •   </os>
  •   <features>
  •     <acpi/>
  •   </features>
  •   <cpu mode='host-passthrough'>
  •     <topology sockets='1' cores='$vm_vcpus' threads='1'/>
  •   </cpu>
  •   <clock offset='variable' adjustment='0' basis=utc'>
  •     <timer name='pit' tickpolicy='delay'/>
  •     <timer name='rtc' tickpolicy='catchup'/>
  •     <timer name='hpet' present='no'/>
  •   </clock>
  •   <on_poweroff>destroy</on_poweroff>
  •   <on_reboot>restart</on_reboot>
  •   <on_crash>restart</on_crash>
  •   <devices>
  •     <emulator>/usr/bin/qemu-kvm</emulator>
  •     <disk type='block' device='disk'>
  •     <driver name='qemu' type='raw' cache='none' io='native'/>
  •     <source dev='$vm_disk'/>
  •     <target dev='vda' bus='virtio'/>
  •     </disk>
  •     <interface type='hostdev' managed='yes'>
  •       <mac address='$mac'/>
  •     <source>
  •       <address type='pci' domain='0x0000' bus='$bus' slot='$slot' function='$function'/>
  •     </source>
  •     </interface>
  •     <serial type='pty'>
  •       <target port='0'/>
  •     </serial>
  •     <console type='pty'>
  •      <target type='serial' port='0'/>
  •     </console>
  •     <input type='mouse' bus='ps2'/>
  •     <graphics type='vnc' port='-1' autoport='yes' keymap='en-us'/>
  •     <memballoon model='none'/>
  •   </devices>
  •   </domain>
• App, Db, and Web VMs have this additional XML:
•   <memoryBacking>
•     <hugepages/>
•   </memoryBacking>
• VMs are bound to host CPUs via their XML configuration in the following manner:
•   App, Db, Web, and Batch VMs of tiles 025-028, 053-056, 081-084, and 109-112 are bound to NUMA node 0.
•   App, Db, Web, and Batch VMs of tiles 001-004, 029-032, 057-060, and 085-088 are bound to NUMA node 1.
•   App, Db, Web, and Batch VMs of tiles 005-008, 033-036, 060-064, and 089-092 are bound to NUMA node 2.
•   All Infra and Mail VMs are bound to NUMA node 3.
•   App, Db, Web, and Batch VMs of tiles 009-012, 037-040, 065-068, and 093-096 are bound to NUMA node 4.
•   App, Db, Web, and Batch VMs of tiles 013-016, 041-044, 069-072, and 097-100 are bound to NUMA node 5.
•   App, Db, Web, and Batch VMs of tiles 017-020, 045-048, 073-076, and 101-104 are bound to NUMA node 6.
•   App, Db, Web, and Batch VMs of tiles 021-024, 049-052, 077-080, and 105-108 are bound to NUMA node 7.
•   App and Batch VMs of tiles 113 and 115 are bound to NUMA node 2.
•   Db VM of tile 113 is bound to NUMA node 1.
•   App and Batch VMs of tiles 114 and 116 are bound to NUMA node 0.
•   Web VM of tile 113 is bound to NUMA node 4.
•   Web VM of tile 114 is bound to NUMA node 5.
•   Web VM of tile 115 is bound to NUMA node 6.
•   Web VM of tile 116 is bound to NUMA node 7.
• App VMs are reniced 2.
• Db VMs are reniced -15.
• Web VMs are reniced -19.
• Mail VMs are reniced 0.
• Infra VMs are reniced 5.
• Batch VMs are reniced 19.

Web Server VM Notes

• The following boot options were used:
• selinux=0 audit=0 elevator=deadline clocksource=tsc tsc=reliable highres=off watchdog=0
• The following services are disabled:
• abrt-ccpp, abrt-oops, abrtd, atd, auditd, autofs, certmonger, cgconfig, cgred, crond, cups, haldaemon, htcacheclean, httpd, ip6tables, iptables, irqbalance, kdump, mcelogd, mdmonitor, messagebus, netconsole, nfs, nfslock, numad, oddjobd, portreserve, postfix, psacct, quota_nld, rdisc, restorecond, rhnsd, rhsmcertd, rngd, rpcgssd, rpcsvcgssd, saslauthd, smartd, sssd, svnserve, winbind, ypbind
• The source code for the Intel kernel module i40evf version 2.0.22 was download from the official web site, built with the default configuration options, and installed.
• Created /sbin/ifup-local with the commands: ethtool -C eth0 adaptive-rx off adaptive-tx off rx-usecs 2500; ethtool -K eth0 gro on gso on
• root file system options: barrier=0,noatime
• The tuned package was installed and activated with the virtual-guest profile.
• The following items were added in /etc/sysctl.conf:
• fs.file-max = 500000
• net.core.somaxconn = 81920
• net.ipv4.tcp_max_tw_buckets = 1500000
• net.ipv4.conf.default.arp_filter = 1
• net.ipv4.conf.all.arp_filter = 1
• net.ipv4.tcp_tw_reuse=1
• net.ipv4.tcp_tw_recycle = 1
• net.ipv4.tcp_keepalive_time = 120
• net.ipv4.tcp_keepalive_intvl = 30
• net.ipv4.tcp_timestamps = 0
• vm.swappiness = 0
• The following lines were added to /etc/security/limits.conf:
• * soft memlock unlimited
• * hard memlock unlimited
• The following commands were run before the benchmark was started:
• ulimit -n 1048576
• ulimit -u 1000000
• The php-fpm package was installed. /etc/php-fpm.d/www.conf changes:
• user = nginx
• group = nginx
• pm.max_children = 700
• pm.start_servers = 100
• pm.min_spare_servers = 100
• pm.max_spare_servers = 200
• /etc/nginx/nginx.conf changes:
• worker_processes 4;
• worker_cpu_affinity 0001 0001 0010 0010;
• worker_rlimit_nofile 65535;
• #error_log /var/log/nginx/error.log warn;
• use epoll;
• worker_connections 65535;
• multi_accept on;
• accept_mutex off;
• # log_format main '$remote_addr - $remote_user [$time_local] "$request" '
• # '$status $body_bytes_sent "$http_referer" '
• # '"$http_user_agent" "$http_x_forwarded_for"';
• log_format common '$remote_addr - $remote_user [$time_local] $request'
•   '"$status" $body_bytes_sent ';
• access_log /var/log/nginx/access.log.gz common gzip buffer=4m flush=1m;
• sendfile_max_chunk 1m;
• tcp_nopush on;
• tcp_nodelay on;
• keepalive_timeout 120;
• open_file_cache max=204800 inactive=20s;
• open_file_cache_min_uses 1;
• open_file_cache_valid 30s;
• open_file_cache_errors on;
• charset utf-8;
• server_names_hash_bucket_size 128;
• client_header_buffer_size 2k;
• large_client_header_buffers 4 4k;
• /etc/nginx/conf.d/default.conf changes:
• listen 80 backlog=81920;
• location / {
•   root /var/www/html;
•   index index.html index.htm index.php;
• }
• location = /error.log {
•   root /var/log/nginx;
• }
• location ~ .*.(html|htm|gif|jpg|jpeg|bmp|png|ico|txt|js|css)$ {
•   root /var/www/html;
•   expires 100d;
• }
• location ~ ^/images/ {
•   root /var/www/html/support;
•   expires 100d;
• }
• location ~ .php$ {
•   root html;
•   fastcgi_pass 127.0.0.1:9000;
•   fastcgi_index index.php;
•   fastcgi_param SCRIPT_FILENAME /var/www/html$fastcgi_script_name;
•   include fastcgi_params;
• }
• /etc/nginx/conf.d/ssl.conf changes:
• listen 443 ssl backlog=81920;
• ssl on;
• ssl_certificate /etc/pki/tls/certs/localhost.crt;
• ssl_certificate_key /etc/pki/tls/private/localhost.key;
• ssl_protocols SSLv3 TLSv1 TLSv1.1 TLSv1.2;
• ssl_ciphers ALL:!ADH:!EXPORT:!SSLv2:RC4+RSA:+HIGH:+MEDIUM:+LOW;
• ssl_prefer_server_ciphers on;
• location / {
•   root /var/www/html;
•   index index.html index.htm index.php;
• }
• location ~ .*.(html|htm|gif|jpg|jpeg|bmp|png|ico|txt|js|css)$ {
•   root /var/www/html;
•   expires 100d;
• }
• location ~ ^/images/ {
•   root /var/www/html/support;
•   expires 100d;
• }
• location ~ .php$ {
•   root html;
•   fastcgi_pass 127.0.0.1:9000;
•   fastcgi_index index.php;
•   fastcgi_param SCRIPT_FILENAME /var/www/html$fastcgi_script_name;
•   include fastcgi_params;
• }
• Web VM used NFS to access files from infra VM.
• After the Infra VM is booted and its NFS service is up, the Web VM mounts the NFS file system via test automation software.
• The datastore is mounted on /var/log/httpd
• NFS mount options used: noatime,actimeo=14400,nocto

Application Server VM Notes

• The following boot options were used:
• selinux=0 audit=0 elevator=deadline clocksource=tsc tsc=reliable highres=off watchdog=0
• The following services are disabled:
• abrt-ccpp, abrt-oops, abrtd, atd, auditd, autofs, certmonger, cgconfig, cgred, crond, cups, haldaemon, ip6tables, iptables, irqbalance, kdump, mcelogd, mdmonitor, messagebus, netconsole, nfs, nfslock, numad, oddjobd, portreserve, postfix, psacct, quota_nld, rdisc, restorecond, rhnsd, rhsmcertd, rngd, rpcgssd, rpcsvcgssd, saslauthd, smartd, sssd, svnserve, winbind, ypbind
• The source code for the Intel kernel module i40evf version 2.0.22 was download from the official web site, built with the default configuration options, and installed.
• Created /sbin/ifup-local with the commands: ethtool -C eth0 adaptive-rx off adaptive-tx off rx-usecs 250; ethtool -K eth0 gro on gso on
• root file system options: barrier=0
• The tuned package was installed and activated with the virtual-guest profile.
• The following items were added in /etc/sysctl.conf:
• fs.file-max = 500000
• net.core.somaxconn = 8192
• net.ipv4.tcp_max_tw_buckets = 500000
• net.ipv4.conf.default.arp_filter = 1
• net.ipv4.conf.all.arp_filter = 1
• vm.swappiness = 0
• The following lines were added to /etc/security/limits.conf:
• * soft memlock unlimited
• * hard memlock unlimited
• The following commands were run before the benchmark was started:
• ulimit -n 1048576
• ulimit -u 1000000
• There are two WebSphere profiles, AppSrv01 for SPECjAppServer2004 and AppSrv02 for SPEC Emulator.
• Heap size for AppSrv01 has been changed to 2880 MB.
• The following changes were made to the defaults in run.properties:
• runLO=4
• warmUpStyle = 1
• linearWarmUpPercentage = .60

Mail Server VM Notes

• The following boot options were used:
• selinux=0 audit=0 elevator=deadline clocksource=tsc tsc=reliable highres=off watchdog=0
• The following services are disabled:
• abrt-ccpp, abrt-oops, abrtd, atd, auditd, autofs, certmonger, cgconfig, cgred, crond, cups, haldaemon, ip6tables, iptables, irqbalance, kdump, mcelogd, mdmonitor, messagebus, netconsole, nfs, nfslock, numad, oddjobd, portreserve, postfix, psacct, quota_nld, rdisc, restorecond, rhnsd, rhsmcertd, rngd, rpcgssd, rpcsvcgssd, saslauthd, smartd, sssd, svnserve, winbind, ypbind
• The source code for the Intel kernel module i40evf version 2.0.22 was download from the official web site, built with the default configuration options, and installed.
• Created /sbin/ifup-local with the commands: ethtool -C eth0 adaptive-rx off adaptive-tx off rx-usecs 1000; ethtool -K eth0 gro on gso on
• root file system options: barrier=0,noatime,data=writeback
• The tuned package was installed and activated with the virtual-guest profile.
• The following items were added in /etc/sysctl.conf:
• fs.file-max = 500000
• net.core.somaxconn = 8192
• net.ipv4.tcp_max_tw_buckets = 500000
• net.ipv4.conf.default.arp_filter = 1
• net.ipv4.conf.all.arp_filter = 1
• The following lines were added to /etc/security/limits.conf:
• * soft memlock unlimited
• * hard memlock unlimited
• The following commands were run before the benchmark was started:
• ulimit -n 1048576
• ulimit -u 1000000
• Dovecot was compiled with default options, and the following dovecot.conf was used:
• base_dir = /var/run/dovecot/
• ssl = no
• ssl_cert =
• ssl_key =
• disable_plaintext_auth = no
• listen = *
• log_path = /var/log/dovecot.log
• mail_debug = no
• mail_location = dbox:/mailstore/dovecot/maildir/%u
• mdbox_rotate_size = 8192
• maildir_very_dirty_syncs = yes
• mbox_very_dirty_syncs = yes
• mbox_write_locks = fcntl
• mmap_disable = yes
• mail_fsync = never
• passdb {
•   args = scheme=plain /etc/passwd.dovecot
•   driver = passwd-file
• }
• protocols = imap
• service auth {
•   user = root
• }
• service imap-login {
•   inet_listener imap {
•     port = 143
•   }
•   process_min_avail = 10
•   service_count = 0
• }
• service imap {
•   process_limit = 1024
• }
• userdb {
•   driver = passwd
• }
• dict {
• }

Database Server VM Notes

• The following boot options were used:
• selinux=0 audit=0 elevator=deadline clocksource=tsc tsc=reliable highres=off watchdog=0
• The following services are disabled:
• abrt-ccpp, abrt-oops, abrtd, atd, auditd, autofs, certmonger, cgconfig, cgred, crond, cups, haldaemon, htcacheclean, httpd, ip6tables, iptables, irqbalance, kdump, mcelogd, mdmonitor, messagebus, netconsole, nfs, nfslock, numad, oddjobd, portreserve, postfix, psacct, quota_nld, rdisc, restorecond, rhnsd, rhsmcertd, rngd, rpcgssd, rpcsvcgssd, saslauthd, smartd, sssd, svnserve, winbind, ypbind
• The source code for the Intel kernel module i40evf version 2.0.22 was download from the official web site, built with the default configuration options, and installed.
• Created /sbin/ifup-local with the commands: ethtool -C eth0 adaptive-rx off adaptive-tx off rx-usecs 250; ethtool -K eth0 gro on gso on
• root file system options: noatime,barrier=0,data=writeback
• The tuned package was installed and activated with the virtual-guest profile.
• The following items were added in /etc/sysctl.conf:
• fs.file-max = 500000
• kernel.randomize_va_space=0
• kernel.sem = 500 64000 400 512
• net.core.somaxconn = 32384
• net.ipv4.tcp_max_tw_buckets = 1000000
• net.ipv4.conf.default.arp_filter = 1
• net.ipv4.conf.all.arp_filter = 1
• vm.swappiness = 0
• The following lines were added to /etc/security/limits.conf:
• * soft memlock unlimited
• * hard memlock unlimited
• The following commands were run before the benchmark was started:
• ulimit -n 1048576
• ulimit -u 1000000
• 13312 2MB huge-pages were allocated prior to starting the database instance.
• Database tuning options:
• db2 update db cfg for specdb using DATABASE_MEMORY 6553600
• db2set DB2_LARGE_PAGE_MEM=DB
• db2set DB2_PINNED_BP=ON
• alter bufferpool IBMDEFAULTBP size 5898240
• alter tablespace roll_space NO FILE SYSTEM CACHING
• alter tablespace C_cust_space NO FILE SYSTEM CACHING
• alter tablespace C_supp_space NO FILE SYSTEM CACHING
• alter tablespace C_site_space NO FILE SYSTEM CACHING
• alter tablespace C_parts_space NO FILE SYSTEM CACHING
• alter tablespace C_custinv_space NO FILE SYSTEM CACHING
• alter tablespace M_parts_space NO FILE SYSTEM CACHING
• alter tablespace M_bom_space NO FILE SYSTEM CACHING
• alter tablespace M_wo_space NO FILE SYSTEM CACHING
• alter tablespace M_lo_space NO FILE SYSTEM CACHING
• alter tablespace M_inv_space NO FILE SYSTEM CACHING
• alter tablespace O_cust_space NO FILE SYSTEM CACHING
• alter tablespace O_ords_space NO FILE SYSTEM CACHING
• alter tablespace O_ordl_space NO FILE SYSTEM CACHING
• alter tablespace O_item_space NO FILE SYSTEM CACHING
• alter tablespace S_comp_space NO FILE SYSTEM CACHING
• alter tablespace S_sc_space NO FILE SYSTEM CACHING
• alter tablespace S_po_space NO FILE SYSTEM CACHING
• alter tablespace S_site_space NO FILE SYSTEM CACHING
• alter tablespace S_supp_space NO FILE SYSTEM CACHING
• alter tablespace U_space NO FILE SYSTEM CACHING
• update database config for specdb using LOGFILSIZ 65536
• update db cfg for specdb using logprimary 64 logsecond 4

Infraserver VM Notes

• The following boot options were used:
• selinux=0 audit=0 elevator=deadline clocksource=tsc tsc=reliable highres=off watchdog=0
• The following services are disabled:
• abrt-ccpp, abrtd, atd, auditd, autofs, certmonger, crond, cups, haldaemon, htcacheclean, httpd, ip6tables, iptables, irqbalance, kdump, mdmonitor, messagebus, netconsole, oddjobd, portreserve, postfix, psacct, quota_nld, rdisc, restorecond, rhnsd, rhsmcertd, rngd, rpcgssd, rpcsvcgssd, saslauthd, smartd, sssd, winbind, ypbind
• The source code for the Intel kernel module i40evf version 2.0.22 was download from the official web site, built with the default configuration options, and installed.
• Created /sbin/ifup-local with the commands: ethtool -C eth0 rx-usecs 333; ethtool -K eth0 gro on gso on
• root file system options: noatime,barrier=0,data=writeback
• The tuned package was installed and activated with the virtual-guest profile.
• The following items were added in /etc/sysctl.conf:
• fs.file-max = 500000
• net.core.somaxconn = 81920
• net.ipv4.tcp_max_tw_buckets = 1500000
• net.ipv4.conf.default.arp_filter = 1
• net.ipv4.conf.all.arp_filter = 1
• net.ipv4.tcp_tw_reuse=1
• net.ipv4.tcp_keepalive_time = 120
• net.ipv4.tcp_keepalive_intvl = 30
• net.ipv4.tcp_timestamps = 0
• net.ipv4.tcp_syn_retries = 4
• net.ipv4.tcp_retries2 = 7
• vm.swappiness=0
• The following lines were added to /etc/security/limits.conf:
• * soft memlock unlimited
• * hard memlock unlimited
• The following commands were run before the benchmark was started:
• ulimit -n 1048576
• ulimit -u 1000000
• httpd.conf changes:
• ServerLimit 10240
• MaxClients 10240
• Listen 81
• fcgi package was built from Fedora 14 source RPM, which includes patch to compile with GCC 4.4.
• The following items were added to/etc/httpd/conf/fcgi.conf:
• LoadModule fastcgi_module modules/mod_fastcgi.so
• ScriptAlias /fcgi-bin/ "/var/www/fcgi-bin/"
• FastCgiIpcDir /etc/httpd/fastcgi
• <Directory "/var/www/fcgi-bin">
•   AllowOverride None
•   Options +ExecCGI -Includes
•   SetHandler fastcgi-script
•   Order allow,deny
•   Allow from all
• </Directory>
• AddHandler fastcgi-script fcgi
• FastCgiServer /var/www/fcgi-bin/besim_fcgi.fcgi -processes 50
• Standard Linux NFS server is enabled to serve files to Web VM.
• /etc/exports file contains:
• /support-download *(rw,async,no_subtree_check,crossmnt,no_root_squash)

Batch Server VM Notes

• The following boot options were used:
• selinux=0 audit=0 elevator=deadline clocksource=tsc tsc=reliable highres=off watchdog=0
• The following services are disabled:
• abrt-ccpp, abrt-oops, abrtd, atd, auditd, autofs, certmonger, cgconfig, cgred, crond, cups, haldaemon, ip6tables, iptables, irqbalance, kdump, mcelogd, mdmonitor, messagebus, netconsole, nfs, nfslock, numad, oddjobd, portreserve, postfix, psacct, quota_nld, rdisc, restorecond, rhnsd, rhsmcertd, rngd, rpcgssd, rpcsvcgssd, saslauthd, smartd, sssd, svnserve, winbind, ypbind
• The source code for the Intel kernel module i40evf version 2.0.22 was download from the official web site, built with the default configuration options, and installed.
• root file system options: barrier=0
• The tuned package was installed and activated with the virtual-guest profile.
• The following items were added in /etc/sysctl.conf:
• fs.file-max = 500000
• net.core.somaxconn = 8192
• net.ipv4.tcp_max_tw_buckets = 500000
• net.ipv4.conf.default.arp_filter = 1
• net.ipv4.conf.all.arp_filter = 1
• The following lines were added to /etc/security/limits.conf:
• * soft memlock unlimited
• * hard memlock unlimited
• The following commands were run before the benchmark was started:
• ulimit -n 1048576
• ulimit -u 1000000

Client Driver Notes

• Eight client systems each host thirteen prime client VMs. One client system hosts twelve prime client VMs. One client system hosts the prime controller VM.
• Each prime client VM has 6 VCPUs, 9 GB memory, and a 16 GB disk. 1 ixgbevf and 1 virtio_blk device are used in all client VMs.
• All prime client VMs run Red Hat Enterprise Linux 6.6.

Power Configuration Notes

• The SUT has four power distribution units (PDUs) each with its own power cable.
• The two power meters were each connected to two of the SUT's power cables.
• There is no power meter for external storage because there is no external storage.

Other Notes

• An external 10Gb Lenovo RackSwitch G8264 switch was used to connect the client systems to the system under test (SUT).
• The SUT used the four dual-port 10Gb PCIe NICs for communication with the VMs.
• The SUT NICs and the external client systems are connected to the switch using 10Gb SFP+ DAC cables.
• An external 1Gb Lenovo RackSwitch G8052 switch was used to manage all the systems.
• All systems are connected to the 1Gb switch via a separate cable plugged into a 1Gb port on the system.

Performance Details: