付録L.3　監視対象がKVMの場合

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

CPU ID (CPU_ID)

「/proc/cpuinfo」ファイルのprocessor行

CPU Name (CPU_NAME)

「/proc/cpuinfo」ファイルのmodel name行

Sampling Time (SAMPLING_TIME)

date -Iseconds

Clocks (CLOCKS)

「/sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq」ファイルの値 / 1000

「/sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq」ファイルが存在しない場合は，「/proc/cpuinfo」ファイルの"cpu MHz"の値

Used (USED)

CLOCKS * (USED_PERCENT / 100)

Unused (UNUSED)

CLOCKS - USED

Used % (USED_PERCENT)

Δ「/proc/stat」ファイルのcpu0〜行，usr+nice+sys列 / Δ収集時間

Unused % (UNUSED_PERCENT)

100 - USED_PERCENT

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

Disk ID (DISK_ID)

「df -lkP」のFileSystem列

Sampling Time (SAMPLING_TIME)

date -Iseconds

Size (SIZE)

「df -lkP」の1024-blocks列 / 1024

Used (USED)

「df -lkP」のUsed列 / 1024

Free (FREE)

「df -lkP」のAvailable列 / 1024

Used % (USED_PERCENT)

(USED / SIZE) * 100

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

Size (SIZE)

「free -m」のMem:行，total列

Sampling Time (SAMPLING_TIME)

date -Iseconds

Used (USED)

「free -m」の-/+ buffers/cache行, used列

VMM Used (VMM_USED)

「ps aux」のRSS列 / 1024

ksmd, virt-manager, libvirtdプロセスの合計

VM Used (VM_USED)

PI_VMI.RESOURCE_USED の合計

Unused (UNUSED)

「free -m」の-/+ buffers/cache行，free列

VM Swap Used (VM_SWAP_USED)

PI_VMI.VM_SWAP_USED の合計

Host Swap Used (HOST_SWAP_USED)

PI_VMI.HOST_SWAP_USED の合計

Total Used (TOTAL_USED)

USED + VM_SWAP_USED + HOST_SWAP_USED

Used % (USED_PERCENT)

(USED / SIZE) * 100

VMM Used % (VMM_USED_PERCENT)

(VMM_USED / SIZE) * 100

VM Used % (VM_USED_PERCENT)

(VM_USED / SIZE) * 100

VM Swap Used % (VM_SWAP_USED_PERCENT)

(VM_SWAP_USED / SIZE) * 100

Host Swap Used % (HOST_SWAP_USED_PERCENT)

(HOST_SWAP_USED / SIZE) * 100

Total Used % (TOTAL_USED_PERCENT)

(TOTAL_USED / SIZE) * 100

Swap IO (SWAP_IO)

SWAP_IN_IO + SWAP_OUT_IO

Swap In IO (SWAP_IN_IO)

「vmstat -s」のpages swapped in 行 * 「getconf PAGE_SIZE」 / 1024 / 1024

Swap Out IO (SWAP_OUT_IO)

「vmstat -s」のpages swapped out 行 * 「getconf PAGE_SIZE」 / 1024 / 1024

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

Net ID (NET_ID)

「ifconfig」のインタフェースのうち，「virsh dumpxml {domain}」で使用していないインタフェース

Sampling Time (SAMPLING_TIME)

date -Iseconds

Rate (RATE)

SEND_RATE + RECV_RATE

Send Rate (SEND_RATE)

Δ「ifconfig」のTX bytes値 / Δ収集時間 / 1024

Recv Rate (RECV_RATE)

Δ「ifconfig」のRX bytes値 / Δ収集時間 / 1024

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

Disk ID (DISK_ID)

「iostat -x -k -d 1 1」のDevice:列

Sampling Time (SAMPLING_TIME)

date -Iseconds

Speed (SPEED)

READ_SPEED + WRITE_SPEED

Read Speed (READ_SPEED)

「iostat -x -k -d 1 1」のrkB/s列

Write Speed (WRITE_SPEED)

「iostat -x -k -d 1 1」のwkB/s列

Requests (REQUESTS)

READ_REQUESTS + WRITE_REQUESTS

Read Requests (READ_REQUESTS)

「iostat -x -k -d 1 1」のr/s列

Write Requests (WRITE_REQUESTS)

「iostat -x -k -d 1 1」のw/s列

Commands (COMMANDS)

−

Abort Commands (ABORT_COMMANDS)

−

Abort Commands % (ABORT_COMMANDS_PERCENT)

−

Bus Resets (BUS_RESETS)

−

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

Status (STATUS)

−

Host Name (HOST_NAME)

監視対象設定

VM_Hostパラメータ

Reason (REASON)

−

Product (PRODUCT)

「virsh version」の各値をカンマ区切りで表示

VM Count (VM_COUNT)

「virsh list -all」の行数

VM Active (VM_ACTIVE)

「virsh list -all」のstatusがrunning,idleの行数

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

Clocks (CLOCKS)

PI_HCI.CLOCKS * COUNT

Count (COUNT)

「/proc/cpuinfo」ファイルのCPU数をカウント

Sampling Time (SAMPLING_TIME)

date -Iseconds

Used (USED)

CLOCKS * USED_PERCENT / 100

VMM Used (VMM_USED)

CLOCKS * VMM_USED_PRECENT / 100

VM Used (VM_USED)

CLOCKS * VM_USED_PERCENT / 100

VMM Console Used (VMM_CONSOLE_USED)

CLOCKS * VMM_CONSOLE_USED_PERCENT / 100

VMM Kernel Used (VMM_KERNEL_USED)

CLOCKS * VMM_KERNEL_USED_PERCENT / 100

VMM Others Used (VMM_OTHERS_USED)

CLOCKS * VMM_OTHERS_USED_PERCENT / 100

Unused (UNUSED)

CLOCKS - USED

Used % (USED_PERCENT)

Δ(「/proc/stat」ファイルのCPU行，usr + nice + sys列) / (Δ収集時刻 * 100 * COUNT) * 100

VMM Used % (VMM_USED_PERCENT)

USED_PERCENT - VM_USED_PERCENT

VM Used % (VM_USED_PERCENT)

Δ(「/proc/stat」ファイルのCPU行,guest列) / (Δ収集時刻 * 100 * COUNT) * 100

VMM Console Used % (VMM_CONSOLE_USED_PERCENT)

「ps aux」のvirt-manager行のPID列

Δ(「top -b -n 1」のvirt-managerのPID行のTIME+列)　/　(Δ収集時刻 * COUNT) * 100

VMM Kernel Used % (VMM_KERNEL_USED_PERCENT)

「ps aux」のlibvirtd, ksmd行のPID列

Δ(「top -b -n 1」のlibvirtd, ksmdのPID行のTIME+列) /　(Δ収集時刻 * COUNT) * 100

VMM Others Used % (VMM_OTHERS_USED_PERCENT)

VMM_USED_PERCENT - VMM_CONSOLE_USED_PERCENT - VMM_KERNEL_USED_PERCENT

Unused % (UNUSED_PERCENT)

UNUSED / CLOCKS * 100

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

PD_VM.VM_ID

CPU ID (CPU_ID)

「virsh vcpuinfo {domain}」のVCPU行

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

PD_VM.VM_NAME

Sampling Time (SAMPLING_TIME)

date -Iseconds

Used (USED)

PI_HCI.CLOCKS * USED_PERCENT / 100

Insufficient (INSUFFICIENT)

−

Request (REQUEST)

−

Used % (USED_PERCENT)

Δ「virsh vcpuinfo {domain}」のCPU time行 / Δ収集時刻 * 100

Insufficient % (INSUFFICIENT_PERCENT)

−

Request % (REQUEST_PERCENT)

−

Used Per Request (USED_PER_REQUEST)

−

Insufficient Per Request (INSUFFICIENT_PER_REQUEST)

−

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

−

Disk ID (DISK_ID)

−

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

−

Sampling Time (SAMPLING_TIME)

−

Size (SIZE)

−

Used (USED)

−

Free (FREE)

−

Used % (USED_PERCENT)

−

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

PD_VM.VM_ID

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

PD_VM.VM_NAME

Sampling Time (SAMPLING_TIME)

date -Iseconds

Size (SIZE)

「pmap -x {pid}」のtotal行，Kbytes値 / 1024

Used (USED)

SIZE

Resource Used (RESOURCE_USED)

「pmap -x {pid}」のtotal行，Dirty値 / 1024

VM Swap Used (VM_SWAP_USED)

「pmap -x {pid}」のtotal行，RSS値 / 1024 - Resource Used

Host Swap Used (HOST_SWAP_USED)

SIZE - 「pmap -x {pid}のtotal行，RSS値 / 1024

Unused (UNUSED)

0固定

Used % (USED_PERCENT)

(USED / SIZE) * 100

Resource Used % (RESOURCE_USED_PERCENT)

(RESOURCE_USED / SIZE) * 100

VM Swap Used % (VM_SWAP_USED_PERCENT)

(VM_SWAP_USED / SIZE) * 100

Host Swap Used % (HOST_SWAP_USED_PERCENT)

(HOST_SWAP_USED / SIZE) * 100

VM Swap IO (VM_SWAP_IO)

−

VM Swap In (VM_SWAP_IN)

−

VM Swap Out (VM_SWAP_OUT)

−

Working Size (WORKING_SIZE)

「virsh dominfo {domain}」のUsed memory行 / 1024

Working Size % (WORKING_SIZE_PERCENT)

(WORKING_SIZE / SIZE) * 100

Share (SHARE)

−

Max (MAX)

「virsh dominfo {domain}」のMax memory行

Min (MIN)

−

Expectation (EXPECTATION)

−

Max % (MAX_PERCENT)

MAX / PI_HMI.SIZE * 100

Min % (MIN_PERCENT)

−

Expectation % (EXPECTATION_PERCENT)

−

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

PD_VM.VM_ID

Net ID (NET_ID)

「virsh dumpxml {domain}」のinterface/mac address

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

PD_VM.VM_NAME

Sampling Time (SAMPLING_TIME)

date -Iseconds

Rate (RATE)

SEND_RATE + RECV_RATE

Send Rate (SEND_RATE)

(Δ「ifconfig」のRX bytes値 / Δ収集時刻) / 1024

Recv Rate (RECV_RATE)

(Δ「ifconfig」のTX bytes値 / Δ収集時刻) / 1024

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

PD_VM.VM_ID

Disk ID (DISK_ID)

「dumpxml {domain}」の<disk type='file' device='disk'>の中の「source file=''」の文字列

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

PD_VM.VM_NAME

Sampling Time (SAMPLING_TIME)

date -Iseconds

Speed (SPEED)

READ_SPEED + WRITE_SPEED

Read Speed (READ_SPEED)

(Δ「virsh domblkstat {domain} {device}」のrd_bytes行 / Δ収集時刻) / 1024

Write Speed (WRITE_SPEED)

(Δ「virsh domblkstat {domain} {device}」のwr_bytes行 / Δ収集時刻) / 1024

Requests (REQUESTS)

READ_REQUESTS + WRITE_REQUESTS

Read Requests (READ_REQUESTS)

Δ「virsh domblkstat {domain} {device}」のrd_req行 / Δ収集時刻

Write Requests (WRITE_REQUESTS)

Δ「virsh domblkstat {domain} {device}」のwr_req行 / Δ収集時刻

Commands (COMMANDS)

−

Abort Commands (ABORT_COMMANDS)

−

Abort Commands % (ABORT_COMMANDS_PERCENT)

−

Bus Resets (BUS_RESETS)

−

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

「virsh dominfo {domain}」のUUID行

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

「virsh list -all」のName列

Status (STATUS)

「virsh dominfo {domain}」のstate行

Information (INFORMATION)

「virsh dominfo {domain}」のOS Type行

Record Type (INPUT_RECORD_TYPE)

−

Record Time (RECORD_TIME)

−

Interval (INTERVAL)

−

VA DeviceID (VADEVICEID)

−

VM ID (VM_ID)

PD_VM.VM_ID

Clocks (CLOCKS)

PI_HCI.CLOCKS * 使用可能CPU数

Count (COUNT)

「virsh dominfo {domain}」のCPU(s)行

VM Host Name (VM_HOST_NAME)

−

VM Name (VM_NAME)

PD_VM.VM_NAME

Sampling Time (SAMPLING_TIME)

date -Iseconds

Used (USED)

CLOCKS * USED_PERCENT / 100

Insufficient (INSUFFICIENT)

−

Request (REQUEST)

−

Host Used % (HOST_USED_PERCENT)

USED / PI.CLOCKS * 100

Used % (USED_PERCENT)

「ps aux」のQEMU行のPID欄

Δ「top -b -n 1」のPID行，TIME+列 / (Δ収集時刻 * COUNT)

Insufficient % (INSUFFICIENT_PERCENT)

−

Request % (REQUEST_PERCENT)

−

Used Per Request (USED_PER_REQUEST)

−

Insufficient Per Request (INSUFFICIENT_PER_REQUEST)

−

Affinity (AFFINITY)

「virsh vcpuinfo {domain}」のCPU Affinity列

Share (SHARE)

−

Max (MAX)

CLOCKS

Min (MIN)

−

Expectation (EXPECTATION)

−

Max % (MAX_PERCENT)

100%固定

Min % (MIN_PERCENT)

−

Expectation % (EXPECTATION_PERCENT)

−