使用 grafana explore 可执行 prometheus query

node_uname_info

熵池大小
node_entropy_available_bits

熵池太小 ，随机函数会阻塞

cpu物理核数
count(count(node_cpu_seconds_total) by (cpu))

node_time_seconds 当前系统时间
node_boot_time_seconds 系统启动时间

系统运行时间
node_time_seconds - node_boot_time_seconds


磁盘空间
node_filesystem_avail_bytes
node_filesystem_avail_bytes{device="/dev/vda1",fstype="ext3",instance="localhost:9100",job="node_exporter",mountpoint="/"}


系统负载
node_load1 node_load5 node_load15


系统cpu使用
node_cpu_seconds_total
node_cpu_seconds_total{cpu="0",instance="localhost:9100",job="node_exporter",mode="user"}
mode 
system user iowait idle nice steal irq softirq

user    处于用户态的运行时间，不包含 nice值为负进程。
nice    nice值为负的进程所占用的CPU时间
system  处于核心态的运行时间
idle    除IO等待时间以外的其它等待时间
iowait IO等待时间(since 2.5.41)
irq  硬中断时间(since 2.6.0-test4)
softirq  软中断时间(since 2.6.0-test4)
stealstolen   
which is the time spent in other operating systems when running in a virtualized environment(since 2.6.11)

guest
which is the time spent running a virtual  CPU  for  guest operating systems under the control of the Linux kernel(since 2.6.24)

总的cpu时间totalCpuTime = user + nice + system + idle + iowait + irq + softirq + stealstolen  +  guest
 
1 - avg without(cpu) (rate(node_cpu_seconds_total{mode="idle"}[2m]))

avg without(cpu,mode) (rate(node_cpu_seconds_total[2m]))

 
内存使用
{__name__=~"node_memory_MemTotal_bytes|node_memory_MemAvailable_bytes|node_memory_Buffers_bytes|node_memory_MemFree_bytes|node_memory_Cached_bytes"}

正则匹配 注意使用 =~  

node_memory_MemTotal_bytes
node_memory_MemAvailable_bytes
node_memory_MemFree_bytes
node_memory_Buffers_bytes
node_memory_Cached_bytes

total available free buffer cache

swap使用 
node_memory_SwapTotal_bytes


node_filefd_allocated  分配的文件句柄数 
node_context_switches_total  上下文切换次数
node_filefd_maximum   最大文件句柄数

网络流量
node_network_receive_bytes_total
node_network_transmit_bytes_total

node_network_receive_bytes_total{device="eth0",instance="localhost:9100",job="node_exporter"}

device 
eth0 lo docker0 vethXXXX

irate(node_network_receive_bytes_total{instance=~'$node',device!~'tap.*|veth.*|br.*|docker.*|virbr*|lo*'}[30m])*8

正则不匹配 !~
正则匹配 =~

网络连接信息
node_netstat_Tcp_CurrEstab
node_sockstat_TCP_tw
node_sockstat_sockets_used
node_sockstat_UDP_inuse
node_sockstat_TCP_alloc


磁盘io

node_disk_reads_completed_total
node_disk_writes_completed_total

node_disk_read_bytes_total
node_disk_written_bytes_total

每次IO读写的耗时
irate(node_disk_read_time_seconds_total{instance=~"$node"}[30m]) / irate(node_disk_reads_completed_total{instance=~"$node"}[30m])
irate(node_disk_write_time_seconds_total{instance=~"$node"}[30m]) / irate(node_disk_writes_completed_total{instance=~"$node"}[30m])


 使用node-exporter监控主机信息