metrics.go 2.8 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394
  1. // Copyright 2015 CoreOS, Inc.
  2. //
  3. // Licensed under the Apache License, Version 2.0 (the "License");
  4. // you may not use this file except in compliance with the License.
  5. // You may obtain a copy of the License at
  6. //
  7. // http://www.apache.org/licenses/LICENSE-2.0
  8. //
  9. // Unless required by applicable law or agreed to in writing, software
  10. // distributed under the License is distributed on an "AS IS" BASIS,
  11. // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  12. // See the License for the specific language governing permissions and
  13. // limitations under the License.
  14. package etcdserver
  15. import (
  16. "time"
  17. "github.com/coreos/etcd/pkg/runtime"
  18. "github.com/prometheus/client_golang/prometheus"
  19. )
  20. var (
  21. // TODO: with label in v3?
  22. proposeDurations = prometheus.NewHistogram(prometheus.HistogramOpts{
  23. Namespace: "etcd_debugging",
  24. Subsystem: "server",
  25. Name: "proposal_duration_seconds",
  26. Help: "The latency distributions of committing proposal.",
  27. Buckets: prometheus.ExponentialBuckets(0.001, 2, 14),
  28. })
  29. proposePending = prometheus.NewGauge(prometheus.GaugeOpts{
  30. Namespace: "etcd_debugging",
  31. Subsystem: "server",
  32. Name: "proposals_pending",
  33. Help: "The current number of pending proposals.",
  34. })
  35. // This is number of proposal failed in client's view.
  36. // The proposal might be later got committed in raft.
  37. proposeFailed = prometheus.NewCounter(prometheus.CounterOpts{
  38. Namespace: "etcd_debugging",
  39. Subsystem: "server",
  40. Name: "proposals_failed_total",
  41. Help: "The total number of failed proposals.",
  42. })
  43. // stable metrics for monitoring
  44. hasLeader = prometheus.NewGauge(prometheus.GaugeOpts{
  45. Namespace: "etcd",
  46. Subsystem: "server",
  47. Name: "has_leader",
  48. Help: "Whether or not a leader exists. 1 is existence, 0 is not.",
  49. })
  50. leaderChanges = prometheus.NewCounter(prometheus.CounterOpts{
  51. Namespace: "etcd",
  52. Subsystem: "server",
  53. Name: "leader_changes_seen_total",
  54. Help: "The number of leader changes seen.",
  55. })
  56. )
  57. func init() {
  58. prometheus.MustRegister(proposeDurations)
  59. prometheus.MustRegister(proposePending)
  60. prometheus.MustRegister(proposeFailed)
  61. prometheus.MustRegister(hasLeader)
  62. prometheus.MustRegister(leaderChanges)
  63. }
  64. func monitorFileDescriptor(done <-chan struct{}) {
  65. ticker := time.NewTicker(5 * time.Second)
  66. defer ticker.Stop()
  67. for {
  68. used, err := runtime.FDUsage()
  69. if err != nil {
  70. plog.Errorf("cannot monitor file descriptor usage (%v)", err)
  71. return
  72. }
  73. limit, err := runtime.FDLimit()
  74. if err != nil {
  75. plog.Errorf("cannot monitor file descriptor usage (%v)", err)
  76. return
  77. }
  78. if used >= limit/5*4 {
  79. plog.Warningf("80%% of the file descriptor limit is used [used = %d, limit = %d]", used, limit)
  80. }
  81. select {
  82. case <-ticker.C:
  83. case <-done:
  84. return
  85. }
  86. }
  87. }