Skip to content

feat: expose sysfs_device label for network, nvme, and infiniband #3746

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sulaimantok wants to merge 1 commit into
base: master
Choose a base branch
from
Open

feat: expose sysfs_device label for network, nvme, and infiniband #3746

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 5 additions & 5 deletions collector/fixtures/e2e-64k-page-output.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1711,8 +1711,8 @@ node_hwmon_temp_max_celsius{chip="platform_coretemp_1",sensor="temp4"} 84
node_hwmon_temp_max_celsius{chip="platform_coretemp_1",sensor="temp5"} 84
# HELP node_infiniband_info Non-numeric data from /sys/class/infiniband/<device>, value is always 1.
# TYPE node_infiniband_info gauge
node_infiniband_info{board_id="I40IW Board ID",device="i40iw0",firmware_version="0.2",hca_type="I40IW"} 1
node_infiniband_info{board_id="SM_1141000001000",device="mlx4_0",firmware_version="2.31.5050",hca_type="MT4099"} 1
node_infiniband_info{board_id="I40IW Board ID",device="i40iw0",firmware_version="0.2",hca_type="I40IW",sysfs_device=""} 1
node_infiniband_info{board_id="SM_1141000001000",device="mlx4_0",firmware_version="2.31.5050",hca_type="MT4099",sysfs_device=""} 1
# HELP node_infiniband_legacy_data_received_bytes_total Number of data octets received on all links
# TYPE node_infiniband_legacy_data_received_bytes_total counter
node_infiniband_legacy_data_received_bytes_total{device="mlx4_0",port="1"} 1.8527668e+07
Expand Down Expand Up @@ -3141,8 +3141,8 @@ node_network_iface_link_mode{device="bond0"} 1
node_network_iface_link_mode{device="eth0"} 1
# HELP node_network_info Non-numeric data from /sys/class/net/<iface>, value is always 1.
# TYPE node_network_info gauge
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="bond0",duplex="full",ifalias="",operstate="up"} 1
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="eth0",duplex="full",ifalias="",operstate="up"} 1
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="bond0",duplex="full",ifalias="",operstate="up",sysfs_device=""} 1
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="eth0",duplex="full",ifalias="",operstate="up",sysfs_device=""} 1
# HELP node_network_mtu_bytes Network device property: mtu_bytes
# TYPE node_network_mtu_bytes gauge
node_network_mtu_bytes{device="bond0"} 1500
Expand Down Expand Up @@ -3484,7 +3484,7 @@ node_nfsd_server_rpcs_total 18628
node_nfsd_server_threads 8
# HELP node_nvme_info Non-numeric data from /sys/class/nvme/<device>, value is always 1.
# TYPE node_nvme_info gauge
node_nvme_info{cntlid="1997",device="nvme0",firmware_revision="1B2QEXP7",model="Samsung SSD 970 PRO 512GB",serial="S680HF8N190894I",state="live"} 1
node_nvme_info{cntlid="1997",device="nvme0",firmware_revision="1B2QEXP7",model="Samsung SSD 970 PRO 512GB",serial="S680HF8N190894I",state="live",sysfs_device=""} 1
# HELP node_nvme_namespace_capacity_bytes Capacity of the NVMe namespace in bytes. Computed as namespace_size * namespace_logical_block_size
# TYPE node_nvme_namespace_capacity_bytes gauge
node_nvme_namespace_capacity_bytes{device="nvme0",nsid="1"} 1.6e+13
Expand Down
10 changes: 5 additions & 5 deletions collector/fixtures/e2e-output.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1743,8 +1743,8 @@ node_hwmon_temp_max_celsius{chip="platform_coretemp_1",sensor="temp4"} 84
node_hwmon_temp_max_celsius{chip="platform_coretemp_1",sensor="temp5"} 84
# HELP node_infiniband_info Non-numeric data from /sys/class/infiniband/<device>, value is always 1.
# TYPE node_infiniband_info gauge
node_infiniband_info{board_id="I40IW Board ID",device="i40iw0",firmware_version="0.2",hca_type="I40IW"} 1
node_infiniband_info{board_id="SM_1141000001000",device="mlx4_0",firmware_version="2.31.5050",hca_type="MT4099"} 1
node_infiniband_info{board_id="I40IW Board ID",device="i40iw0",firmware_version="0.2",hca_type="I40IW",sysfs_device=""} 1
node_infiniband_info{board_id="SM_1141000001000",device="mlx4_0",firmware_version="2.31.5050",hca_type="MT4099",sysfs_device=""} 1
# HELP node_infiniband_legacy_data_received_bytes_total Number of data octets received on all links
# TYPE node_infiniband_legacy_data_received_bytes_total counter
node_infiniband_legacy_data_received_bytes_total{device="mlx4_0",port="1"} 1.8527668e+07
Expand Down Expand Up @@ -3173,8 +3173,8 @@ node_network_iface_link_mode{device="bond0"} 1
node_network_iface_link_mode{device="eth0"} 1
# HELP node_network_info Non-numeric data from /sys/class/net/<iface>, value is always 1.
# TYPE node_network_info gauge
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="bond0",duplex="full",ifalias="",operstate="up"} 1
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="eth0",duplex="full",ifalias="",operstate="up"} 1
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="bond0",duplex="full",ifalias="",operstate="up",sysfs_device=""} 1
node_network_info{address="01:01:01:01:01:01",adminstate="up",broadcast="ff:ff:ff:ff:ff:ff",device="eth0",duplex="full",ifalias="",operstate="up",sysfs_device=""} 1
# HELP node_network_mtu_bytes Network device property: mtu_bytes
# TYPE node_network_mtu_bytes gauge
node_network_mtu_bytes{device="bond0"} 1500
Expand Down Expand Up @@ -3516,7 +3516,7 @@ node_nfsd_server_rpcs_total 18628
node_nfsd_server_threads 8
# HELP node_nvme_info Non-numeric data from /sys/class/nvme/<device>, value is always 1.
# TYPE node_nvme_info gauge
node_nvme_info{cntlid="1997",device="nvme0",firmware_revision="1B2QEXP7",model="Samsung SSD 970 PRO 512GB",serial="S680HF8N190894I",state="live"} 1
node_nvme_info{cntlid="1997",device="nvme0",firmware_revision="1B2QEXP7",model="Samsung SSD 970 PRO 512GB",serial="S680HF8N190894I",state="live",sysfs_device=""} 1
# HELP node_nvme_namespace_capacity_bytes Capacity of the NVMe namespace in bytes. Computed as namespace_size * namespace_logical_block_size
# TYPE node_nvme_namespace_capacity_bytes gauge
node_nvme_namespace_capacity_bytes{device="nvme0",nsid="1"} 1.6e+13
Expand Down
11 changes: 11 additions & 0 deletions collector/helper.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,11 +15,22 @@ package collector

import (
"os"
"path/filepath"
"regexp"
"strconv"
"strings"
)

// sysfsDevice returns the base name of the symlink target of /sys/class/<class>/<device>/device
func sysfsDevice(class, device string) string {
path := sysFilePath(filepath.Join("class", class, device, "device"))
symlink, err := os.Readlink(path)
if err != nil {
return ""
}
return filepath.Base(symlink)
}

func readUintFromFile(path string) (uint64, error) {
data, err := os.ReadFile(path)
if err != nil {
Expand Down
4 changes: 2 additions & 2 deletions collector/infiniband_linux.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -175,11 +175,11 @@ func (c *infinibandCollector) Update(ch chan<- prometheus.Metric) error {
infoDesc := prometheus.NewDesc(
prometheus.BuildFQName(namespace, c.subsystem, "info"),
"Non-numeric data from /sys/class/infiniband/<device>, value is always 1.",
[]string{"device", "board_id", "firmware_version", "hca_type"},
[]string{"device", "board_id", "firmware_version", "hca_type", "sysfs_device"},
nil,
)
infoValue := 1.0
ch <- prometheus.MustNewConstMetric(infoDesc, prometheus.GaugeValue, infoValue, device.Name, device.BoardID, device.FirmwareVersion, device.HCAType)
ch <- prometheus.MustNewConstMetric(infoDesc, prometheus.GaugeValue, infoValue, device.Name, device.BoardID, device.FirmwareVersion, device.HCAType, sysfsDevice("infiniband", device.Name))

for _, port := range device.Ports {
portStr := strconv.FormatUint(uint64(port.Port), 10)
Expand Down
4 changes: 2 additions & 2 deletions collector/netclass_linux.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -97,12 +97,12 @@ func (c *netClassCollector) netClassSysfsUpdate(ch chan<- prometheus.Metric) err
infoDesc := prometheus.NewDesc(
prometheus.BuildFQName(namespace, c.subsystem, "info"),
"Non-numeric data from /sys/class/net/<iface>, value is always 1.",
[]string{"device", "address", "broadcast", "duplex", "operstate", "adminstate", "ifalias"},
[]string{"device", "address", "broadcast", "duplex", "operstate", "adminstate", "ifalias", "sysfs_device"},
nil,
)
infoValue := 1.0

ch <- prometheus.MustNewConstMetric(infoDesc, prometheus.GaugeValue, infoValue, ifaceInfo.Name, ifaceInfo.Address, ifaceInfo.Broadcast, ifaceInfo.Duplex, ifaceInfo.OperState, getAdminState(ifaceInfo.Flags), ifaceInfo.IfAlias)
ch <- prometheus.MustNewConstMetric(infoDesc, prometheus.GaugeValue, infoValue, ifaceInfo.Name, ifaceInfo.Address, ifaceInfo.Broadcast, ifaceInfo.Duplex, ifaceInfo.OperState, getAdminState(ifaceInfo.Flags), ifaceInfo.IfAlias, sysfsDevice("net", ifaceInfo.Name))

pushMetric(ch, c.getFieldDesc("address_assign_type"), ifaceInfo.AddrAssignType, prometheus.GaugeValue, ifaceInfo.Name)
pushMetric(ch, c.getFieldDesc("carrier"), ifaceInfo.Carrier, prometheus.GaugeValue, ifaceInfo.Name)
Expand Down
3 changes: 2 additions & 1 deletion collector/nvme_linux.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ var (
nvmeInfo = prometheus.NewDesc(
prometheus.BuildFQName(namespace, "nvme", "info"),
"Non-numeric data from /sys/class/nvme/<device>, value is always 1.",
[]string{"device", "firmware_revision", "model", "serial", "state", "cntlid"},
[]string{"device", "firmware_revision", "model", "serial", "state", "cntlid", "sysfs_device"},
nil,
)
nvmeNamespaceInfo = prometheus.NewDesc(
Expand Down Expand Up @@ -106,6 +106,7 @@ func (c *nvmeCollector) Update(ch chan<- prometheus.Metric) error {
device.Serial,
device.State,
device.ControllerID,
sysfsDevice("nvme", device.Name),
)

// Export namespace-level metrics
Expand Down