ston1th/node_exporter
0
0
Fork 0
Code Issues Releases Activity

Address first batch of old review comments

Browse source 
Signed-off-by: beorn7 <beorn@grafana.com>
This commit is contained in:
beorn7 2019-07-12 22:58:43 +02:00
commit 2180c2f3bf
3 changed files with 105 additions and 65 deletions
Download patch file Download diff file Expand all files Collapse all files

49
docs/node-mixin/rules/rules.libsonnet
View file

@ -8,8 +8,8 @@
// This rule gives the number of CPUs per node.
record: 'instance:node_num_cpu:sum',
expr: |||
count by (instance) (
sum by (instance, cpu) (
count without (cpu) (
sum without (mode) (
node_cpu_seconds_total{%(nodeExporterSelector)s}
)
)
@ -19,29 +19,20 @@
// CPU utilisation is % CPU is not idle.
record: 'instance:node_cpu_utilisation:avg1m',
expr: |||
1 - avg by (instance) (
rate(node_cpu_seconds_total{%(nodeExporterSelector)s,mode="idle"}[1m])
1 - avg without (cpu, mode) (
rate(node_cpu_seconds_total{%(nodeExporterSelector)s, mode="idle"}[1m])
)
||| % $._config,
},
{
// CPU saturation is 1min avg run queue length / number of CPUs.
// Can go over 100%. >100% is bad.
record: 'instance:node_cpu_saturation_load1:',
// This is CPU saturation: 1min avg run queue length / number of CPUs.
// Can go over 1. >1 is bad.
record: 'instance:node_load1_per_cpu:ratio',
expr: |||
(
sum by (instance) (node_load1{%(nodeExporterSelector)s})
node_load1{%(nodeExporterSelector)s}
/
instance:node_num_cpu:sum
)
||| % $._config,
},
{
// Total memory per node
record: 'instance:node_memory_bytes_total:sum',
expr: |||
sum by (instance) (
node_memory_MemTotal_bytes{%(nodeExporterSelector)s}
instance:node_num_cpu:sum{%(nodeExporterSelector)s}
)
||| % $._config,
},
@ -57,9 +48,9 @@
||| % $._config,
},
{
record: 'instance:node_memory_swap_io_bytes:sum_rate',
record: 'instance:node_memory_swap_io_pages:sum_rate',
expr: |||
1e3 * sum by (instance) (
(
rate(node_vmstat_pgpgin{%(nodeExporterSelector)s}[1m])
+
rate(node_vmstat_pgpgout{%(nodeExporterSelector)s}[1m])
@ -70,7 +61,7 @@
// Disk utilisation (ms spent, 1 second irate())
record: 'instance:node_disk_utilisation:sum_irate',
expr: |||
sum by (instance) (
sum without (device) (
irate(node_disk_io_time_seconds_total{%(nodeExporterSelector)s,device=~"(sd|xvd).+"}[1m])
)
||| % $._config,
@ -79,28 +70,30 @@
// Disk saturation (ms spent, by rate() it's bound by 1 second)
record: 'instance:node_disk_saturation:sum_irate',
expr: |||
sum by (instance) (
sum without (device) (
irate(node_disk_io_time_weighted_seconds_total{%(nodeExporterSelector)s,device=~"(sd|xvd).+"}[1m])
)
||| % $._config,
},
// TODO: For the following two rules, consider configurable filtering to exclude more network
// device names than just "lo".
{
record: 'instance:node_net_utilisation:sum_irate',
expr: |||
sum by (instance) (
irate(node_network_receive_bytes_total{%(nodeExporterSelector)s,device=~"eth[0-9]+"}[1m])
sum without (device) (
irate(node_network_receive_bytes_total{%(nodeExporterSelector)s, device!="lo"}[1m])
+
irate(node_network_transmit_bytes_total{%(nodeExporterSelector)s,device=~"eth[0-9]+"}[1m])
irate(node_network_transmit_bytes_total{%(nodeExporterSelector)s, device!="lo"}[1m])
)
||| % $._config,
},
{
record: 'instance:node_net_saturation:sum_irate',
expr: |||
sum by (instance) (
irate(node_network_receive_drop_total{%(nodeExporterSelector)s,device=~"eth[0-9]+"}[1m])
sum without (device) (
irate(node_network_receive_drop_total{%(nodeExporterSelector)s, device!="lo"}[1m])
+
irate(node_network_transmit_drop_total{%(nodeExporterSelector)s,device=~"eth[0-9]+"}[1m])
irate(node_network_transmit_drop_total{%(nodeExporterSelector)s, device!="lo"}[1m])
)
||| % $._config,
},