prometheus告警指标（必威体育精装版文档）.pdf

prometheus告告警警指指标标

记记录录了了prometheus告告警警指指标标

主主机机和和硬硬件件监监控控

可可⽤⽤内内存存指指标标

主机中可⽤内存容量不⾜10%

-alert:HostOutOfMemory

expr:node_memory_MemAvailable_bytes/node_memory_MemTotal_bytes*10010

for:5m

labels:

severity:warning

annotations:

summary:Hostoutofmemory(instance{{$labels.instance}})

description:Nodememoryisfillingup(10%left)\nVALUE={{$value}}\nLABELS:{{$labels}}

内内存存

节点内存压⼒。主要页⾯故障率⾼

-alert:HostMemoryUnderMemoryPressure

expr:rate(node_vmstat_pgmajfault[1m])1000

for:5m

labels:

severity:warning

annotations:

summary:Hostmemoryundermemorypressure(instance{{$labels.instance}})

description:Thenodeisunderheavymemorypressure.Highrateofmajorpagefaults\nVALUE={{$value}}\nLABELS:{{$labels}}

主主机机⽹⽹络络接接⼝⼝流流⼊⼊流流量量异异常常

主机⽹络接⼝可能接收了太多的数据（100MB/s）。阀值根据⾃⼰机器背板⽹卡决定

-alert:HostUnusualNetworkThroughputIn

expr:sumby(instance)(rate(node_network_receive_bytes_total[2m]))/1024/1024100

for:5m

labels:

severity:warning

annotations:

summary:Hostunusualnetworkthroughputin(instance{{$labels.instance}})

description:Hostnetworkinterfacesareprobablyreceivingtoomuchdata(100MB/s)\nVALUE={{$value}}\nLABELS:{{$labels}}

主主机机⽹⽹络络接接⼝⼝流流出出流流量量异异常常

主机⽹络接⼝可能发送了太多的数据（100MB/s）。

-alert:HostUnusualNetworkThroughputOut

expr:sumby(instance)(rate(node_network_transmit_bytes_total[2m]))/1024/1024100

for:5m

labels:

severity:warning

annotations:

summary:Hostunusualnetworkthroughputout(instance{{$labels.instance}})

descriptio