Elastic GPU Service Monitoring Indicators

Monitoring item Chinese name Unit Tag description

agent.alive

Agent survival

-

-

cpu.busy

CPU usage percentage

%

-

cpu.iowait

CPU IO waiting time

ms

-

cpu.irq

CPU interrupt time

Times

-

cpu.softirq

CPU soft interrupt

Times

-

cpu.switches

CPU context switch

Times

-

cpu.system

CPU system time

%

-

cpu.user

CPU user time

%

-

df.bytes.free

Partition free space size

byte

fstype=, mount=

df.bytes.free.percent

Partition space vacancy percentage

%

fstype=, mount=

df.bytes.total

Partition size

byte

fstype=, mount=

df.bytes.used

Partition differentiated usage

byte

fstype=, mount=

df.bytes.used.percent

Partition usage percentage

%

fstype=, mount=

df.inodes.free.percent

Inode vacancy percentage

%

fstype=, mount=

df.statistics.total

Disk total size

byte

-

df.statistics.used

Total disk usage

byte

-

df.statistics.used.percent

Disk usage percentage

%

-

disk.io.avgqu-sz

Average size of IO waiting queue

-

device=

disk.io.avgrq_sz

Average size of IO request

-

device=

disk.io.await

IO waiting time

ms

device=

disk.io.read_bytes

Number of bytes to read

byte/s

device=

disk.io.read_requests

Number of requests to read

req/s

device=

disk.io.read_sectors

Number of sectors to read

sec/s

device=

disk.io.svctm

Average service time

ms

device=

disk.io.util

Disk load

%

device=

disk.io.write_bytes

Number of bytes to write

byte/s

device=

disk.io.write_requests

Number of requests to write

-

device=

disk.io.write_sectors

Number of sectors to write

-

device=

gpu.fan.speed

CPU fan speed

%

-

gpu.mem.memfree

Free GPU memory space

byte

minor-number=

gpu.mem.memfree.all

Total free GPU memory space

byte

-

gpu.mem.memtotal

Total GPU memory

byte

minor-number=

gpu.mem.memtotal.all

Total GPU size

byte

-

gpu.mem.memused

GPU memory usage

byte

minor-number=

gpu.mem.memused.all

Total GPU memory usage

byte

-

gpu.mem.memused.percent

GPU memory usage ratio

%

minor-number=

gpu.mem.memused.percent.all

GPU memory usage ratio

%

-

gpu.num

GPU number

pcs

-

gpu.perf

GPU performance

-

-

gpu.power.draw

GPU power draw

W

minor-number=

gpu.power.draw.percent

GPU power draw percent

%

minor-number=

gpu.power.limit

GPU power draw limit

W

minor-number=

gpu.temperature

GPU temperature

Degrees celsius

minor-number=

gpu.util

GPU usage ratio

%

minor-number=

gpu.util.all

Total GPU usage ratio

%

-

load.15min

15 minutes load

-

-

load.1min

1 minute load

-

-

load.5min

5 minutes load

-

-

mem.memfree

Free memory space

byte

-

mem.memfree.percent

Memory vacancy percentage

%

-

mem.memtotal

Total memory space

byte

-

mem.memused

Memory usage

byte

-

mem.memused.percent

Memory usage ratio

%

-

mem.swapfree

Free swap space

byte

-

mem.swapfree.percent

Swap vacancy percentage

%

-

mem.swaptotal

Total swap space

byte

-

mem.swapused

Swap usage

byte

-

mem.swapused.percent

Swap usage percentage

%

-

net.if.in.bytes

Inflow

byte/s

iface=

net.if.in.dropped

Packet loss volume of incoming packets

-

iface=

net.if.in.packets

Incoming packet volume

-

iface=

net.if.out.bytes

Outflow

-

iface=

net.if.out.dropped

Packet loss volume of outgoing packets

-

iface=

net.if.out.packets

Outgoing packets volume

-

iface=

net.if.total.bytes

Total data flow

byte/s

-

net.if.total.packets

Total network packet quantity

-

iface=

net.port.listen

Port monitoring

-

port=

ping.available

Ping connectivity

-

-

ping.delay.avg

Average ping delay

-

-

ping.delay.max

Maximum ping delay

-

-

ping.delay.min

Minimum ping delay

-

-

ping.loss

Ping packet loss percentage

-

-

proc.num

Process number

-

cmdline=, name=

ss.closed

Number of CLOSED connection

-

-

ss.estab

Number of ESTABLISH connection

-

-

ss.synrecv

Number of SYN_RECV connection

-

-

ss.timewait

Number of TIME_WAIT connection

-

-

Did the above content solve your problem? Yes No
Please complete information!

Call us

400-151-8800

Email us

cloud@pingan.com

Online customer service

Instant reply

Technical Support

cloud products