feat(inputs.infiniband_hw): add a new input plugin for Infiniband car…

…d/ports HW statistics

plugins/inputs/all/infiniband_hw.go

@@ -0,0 +1,5 @@
+//go:build !custom || inputs || inputs.infiniband_hw
+
+package all
+
+import _ "github.com/influxdata/telegraf/plugins/inputs/infiniband_hw" // register plugin

plugins/inputs/infiniband_hw/README.md

@@ -0,0 +1,75 @@
+# InfiniBand Hardware Input Plugin
+
+This plugin gathers statistics for all InfiniBand devices and ports on the
+system. These are the hardware counters that can be found in
+`/sys/class/infiniband/<dev>/ports/<port>/hw_counters/`
+
+**Supported Platforms**: Linux
+
+## Global configuration options <!-- @/docs/includes/plugin_config.md -->
+
+In addition to the plugin-specific configuration settings, plugins support
+additional global and plugin configuration settings. These settings are used to
+modify metrics, tags, and field or create aliases and configure ordering, etc.
+See the [CONFIGURATION.md][CONFIGURATION.md] for more details.
+
+[CONFIGURATION.md]: ../../../docs/CONFIGURATION.md#plugins
+
+## Configuration
+
+```toml @sample.conf
+# Gets hardware counters from all InfiniBand cards and ports installed
+# This plugin ONLY supports Linux
+[[inputs.infiniband_hw]]
+  # no configuration
+```
+
+## Metrics
+
+Actual metrics depend on the InfiniBand devices, the plugin uses a simple
+mapping from hw_counter -> hw_counter value.
+
+[Information about hw_counters][hw_counters] collected is provided by Nvidia.
+
+[hw_counters]: https://enterprise-support.nvidia.com/s/article/understanding-mlx5-linux-counters-and-status-parameters
+
+- infiniband
+  - tags:
+    - device
+    - port
+  - fields:
+    - duplicate_request (integer)
+    - implied_nak_seq_err (integer)
+    - lifespan (integer)
+    - local_ack_timeout_err (integer)
+    - np_cnp_sent (integer)
+    - np_ecn_marked_roce_packets (integer)
+    - out_of_buffer (integer)
+    - out_of_sequence (integer)
+    - packet_seq_err (integer)
+    - req_cqe_error (integer)
+    - req_cqe_flush_error (integer)
+    - req_remote_access_errors (integer)
+    - req_remote_invalid_request (integer)
+    - resp_cqe_error (integer)
+    - resp_cqe_flush_error (integer)
+    - resp_local_length_error (integer)
+    - resp_remote_access_errors (integer)
+    - rnr_nak_retry_err (integer)
+    - roce_adp_retrans (integer)
+    - roce_adp_retrans_to (integer)
+    - roce_slow_restart (integer)
+    - roce_slow_restart_cnps (integer)
+    - roce_slow_restart_trans (integer)
+    - rp_cnp_handled (integer)
+    - rp_cnp_ignored (integer)
+    - rx_atomic_requests (integer)
+    - rx_icrc_encapsulated (integer)
+    - rx_read_requests (integer)
+    - rx_write_requests (integer)
+
+## Example Output
+
+```text
+infiniband_hw,device=mlx5_4,host=host1,port=1 local_ack_timeout_err=0i,req_cqe_error=0i,roce_slow_restart=0i,roce_adp_retrans=0i,rx_atomic_requests=0i,np_ecn_marked_roce_packets=0i,rp_cnp_handled=0i,req_remote_access_errors=0i,np_cnp_sent=0i,resp_cqe_error=0i,out_of_sequence=0i,roce_slow_restart_cnps=0i,req_remote_invalid_request=0i,implied_nak_seq_err=0i,rp_cnp_ignored=0i,resp_local_length_error=0i,lifespan=10i,out_of_buffer=0i,rx_write_requests=0i,resp_cqe_flush_error=0i,rx_icrc_encapsulated=0i,rx_read_requests=0i,resp_remote_access_errors=0i,roce_adp_retrans_to=0i,roce_slow_restart_trans=0i,rnr_nak_retry_err=0i,req_cqe_flush_error=0i,packet_seq_err=0i,duplicate_request=0i 1734520190000000000
+```

plugins/inputs/infiniband_hw/infiniband_hw.go

@@ -0,0 +1,25 @@
+//go:generate ../../../tools/readme_config_includer/generator
+package infiniband_hw
+
+import (
+	_ "embed"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/inputs"
+)
+
+//go:embed sample.conf
+var sampleConfig string
+
+type InfinibandHW struct {
+	Log telegraf.Logger `toml:"-"`
+}
+
+func (*InfinibandHW) SampleConfig() string {
+	return sampleConfig
+}
+
+// Initialise plugin
+func init() {
+	inputs.Add("infiniband_hw", func() telegraf.Input { return &InfinibandHW{} })
+}

plugins/inputs/infiniband_hw/infiniband_hw_linux.go

@@ -0,0 +1,53 @@
+//go:build linux
+
+package infiniband_hw
+
+import (
+	"errors"
+	"strconv"
+
+	"github.com/Mellanox/rdmamap"
+
+	"github.com/influxdata/telegraf"
+)
+
+// Gather hardware statistics from our infiniband cards
+func (i *InfinibandHW) Gather(acc telegraf.Accumulator) error {
+	rdmaDevices := rdmamap.GetRdmaDeviceList()
+
+	if len(rdmaDevices) == 0 {
+		return errors.New("no InfiniBand devices found in /sys/class/infiniband/")
+	}
+
+	for _, dev := range rdmaDevices {
+		devicePorts := rdmamap.GetPorts(dev)
+		for _, port := range devicePorts {
+			portInt, err := strconv.Atoi(port)
+			if err != nil {
+				return err
+			}
+
+			stats, err := rdmamap.GetRdmaSysfsHwStats(dev, portInt)
+			if err != nil {
+				continue
+			}
+
+			addStats(dev, port, stats, acc)
+		}
+	}
+
+	return nil
+}
+
+// Add the statistics to the accumulator
+func addStats(dev, port string, stats []rdmamap.RdmaStatEntry, acc telegraf.Accumulator) {
+	// Allow users to filter by card and port
+	tags := map[string]string{"device": dev, "port": port}
+	fields := make(map[string]interface{})
+
+	for _, entry := range stats {
+		fields[entry.Name] = entry.Value
+	}
+
+	acc.AddFields("infiniband_hw", fields, tags)
+}

plugins/inputs/infiniband_hw/infiniband_hw_notlinux.go

@@ -0,0 +1,23 @@
+//go:build !linux
+
+package infiniband_hw
+
+import (
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/inputs"
+)
+
+func (i *InfinibandHW) Init() error {
+	i.Log.Warn("Current platform is not supported")
+	return nil
+}
+
+func (*InfinibandHW) Gather(_ telegraf.Accumulator) error {
+	return nil
+}
+
+func init() {
+	inputs.Add("infiniband_hw", func() telegraf.Input {
+		return &InfinibandHW{}
+	})
+}

plugins/inputs/infiniband_hw/infiniband_hw_test.go

@@ -0,0 +1,175 @@
+//go:build linux
+
+package infiniband_hw
+
+import (
+	"testing"
+
+	"github.com/Mellanox/rdmamap"
+
+	"github.com/influxdata/telegraf/testutil"
+)
+
+func TestInfinibandHw(t *testing.T) {
+	fields := map[string]interface{}{
+		"duplicate_request":          uint64(0),
+		"implied_nak_seq_err":        uint64(0),
+		"lifespan":                   uint64(10),
+		"local_ack_timeout_err":      uint64(38),
+		"np_cnp_sent":                uint64(10284520),
+		"np_ecn_marked_roce_packets": uint64(286733949),
+		"out_of_buffer":              uint64(1149772),
+		"out_of_sequence":            uint64(44),
+		"packet_seq_err":             uint64(1),
+		"req_cqe_error":              uint64(10776),
+		"req_cqe_flush_error":        uint64(2173),
+		"req_remote_access_errors":   uint64(0),
+		"req_remote_invalid_request": uint64(0),
+		"resp_cqe_error":             uint64(759),
+		"resp_cqe_flush_error":       uint64(759),
+		"resp_local_length_error":    uint64(0),
+		"resp_remote_access_errors":  uint64(0),
+		"rnr_nak_retry_err":          uint64(0),
+		"roce_adp_retrans":           uint64(0),
+		"roce_adp_retrans_to":        uint64(0),
+		"roce_slow_restart":          uint64(0),
+		"roce_slow_restart_cnps":     uint64(0),
+		"roce_slow_restart_trans":    uint64(0),
+		"rp_cnp_handled":             uint64(1),
+		"rp_cnp_ignored":             uint64(0),
+		"rx_atomic_requests":         uint64(0),
+		"rx_icrc_encapsulated":       uint64(0),
+		"rx_read_requests":           uint64(488228),
+		"rx_write_requests":          uint64(3928699),
+	}
+
+	tags := map[string]string{
+		"device": "m1x5_0",
+		"port":   "1",
+	}
+
+	sampleRdmaHwstatsEntries := []rdmamap.RdmaStatEntry{
+		{
+			Name:  "duplicate_request",
+			Value: uint64(0),
+		},
+		{
+			Name:  "implied_nak_seq_err",
+			Value: uint64(0),
+		},
+		{
+			Name:  "lifespan",
+			Value: uint64(10),
+		},
+		{
+			Name:  "local_ack_timeout_err",
+			Value: uint64(38),
+		},
+		{
+			Name:  "np_cnp_sent",
+			Value: uint64(10284520),
+		},
+		{
+			Name:  "np_ecn_marked_roce_packets",
+			Value: uint64(286733949),
+		},
+		{
+			Name:  "out_of_buffer",
+			Value: uint64(1149772),
+		},
+		{
+			Name:  "out_of_sequence",
+			Value: uint64(44),
+		},
+		{
+			Name:  "packet_seq_err",
+			Value: uint64(1),
+		},
+		{
+			Name:  "req_cqe_error",
+			Value: uint64(10776),
+		},
+		{
+			Name:  "req_cqe_flush_error",
+			Value: uint64(2173),
+		},
+		{
+			Name:  "req_remote_access_errors",
+			Value: uint64(0),
+		},
+		{
+			Name:  "req_remote_invalid_request",
+			Value: uint64(0),
+		},
+		{
+			Name:  "resp_cqe_error",
+			Value: uint64(759),
+		},
+		{
+			Name:  "resp_cqe_flush_error",
+			Value: uint64(759),
+		},
+		{
+			Name:  "resp_local_length_error",
+			Value: uint64(0),
+		},
+		{
+			Name:  "resp_remote_access_errors",
+			Value: uint64(0),
+		},
+		{
+			Name:  "rnr_nak_retry_err",
+			Value: uint64(0),
+		},
+		{
+			Name:  "roce_adp_retrans",
+			Value: uint64(0),
+		},
+		{
+			Name:  "roce_adp_retrans_to",
+			Value: uint64(0),
+		},
+		{
+			Name:  "roce_slow_restart",
+			Value: uint64(0),
+		},
+		{
+			Name:  "roce_slow_restart_cnps",
+			Value: uint64(0),
+		},
+		{
+			Name:  "roce_slow_restart_trans",
+			Value: uint64(0),
+		},
+		{
+			Name:  "rp_cnp_handled",
+			Value: uint64(1),
+		},
+		{
+			Name:  "rp_cnp_ignored",
+			Value: uint64(0),
+		},
+		{
+			Name:  "rx_atomic_requests",
+			Value: uint64(0),
+		},
+		{
+			Name:  "rx_icrc_encapsulated",
+			Value: uint64(0),
+		},
+		{
+			Name:  "rx_read_requests",
+			Value: uint64(488228),
+		},
+		{
+			Name:  "rx_write_requests",
+			Value: uint64(3928699),
+		},
+	}
+
+	var acc testutil.Accumulator
+
+	addStats("m1x5_0", "1", sampleRdmaHwstatsEntries, &acc)
+
+	acc.AssertContainsTaggedFields(t, "infiniband_hw", fields, tags)
+}

plugins/inputs/infiniband_hw/sample.conf

@@ -0,0 +1,4 @@
+# Gets hardware counters from all InfiniBand cards and ports installed
+# This plugin ONLY supports Linux
+[[inputs.infiniband_hw]]
+  # no configuration