github.com/mattyr/nomad@v0.3.3-0.20160919021406-3485a065154a/website/source/docs/agent/telemetry.html.md (about)

     1  ---
     2  layout: "docs"
     3  page_title: "Telemetry"
     4  sidebar_current: "docs-agent-telemetry"
     5  description: |-
     6    Learn about the telemetry data available in Nomad.
     7  ---
     8  
     9  # Telemetry
    10  
    11  The Nomad agent collects various runtime metrics about the performance of
    12  different libraries and subsystems. These metrics are aggregated on a ten
    13  second interval and are retained for one minute.
    14  
    15  To view this data, you must send a signal to the Nomad process: on Unix,
    16  this is `USR1` while on Windows it is `BREAK`. Once Nomad receives the signal,
    17  it will dump the current telemetry information to the agent's `stderr`.
    18  
    19  This telemetry information can be used for debugging or otherwise
    20  getting a better view of what Nomad is doing.
    21  
    22  Telemetry information can be streamed to both [statsite](https://github.com/armon/statsite)
    23  as well as statsd based on providing the appropriate configuration options.
    24  
    25  To configure the telemetry output please see the [agent
    26  configuration](/docs/agent/config.html#telemetry_config).
    27  
    28  Below is sample output of a telemetry dump:
    29  
    30  ```text
    31  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.nomad.broker.total_blocked': 0.000
    32  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.nomad.plan.queue_depth': 0.000
    33  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.malloc_count': 7568.000
    34  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.total_gc_runs': 8.000
    35  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.nomad.broker.total_ready': 0.000
    36  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.num_goroutines': 56.000
    37  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.sys_bytes': 3999992.000
    38  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.heap_objects': 4135.000
    39  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.nomad.heartbeat.active': 1.000
    40  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.nomad.broker.total_unacked': 0.000
    41  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.nomad.broker.total_waiting': 0.000
    42  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.alloc_bytes': 634056.000
    43  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.free_count': 3433.000
    44  [2015-09-17 16:59:40 -0700 PDT][G] 'nomad.runtime.total_gc_pause_ns': 6572135.000
    45  [2015-09-17 16:59:40 -0700 PDT][C] 'nomad.memberlist.msg.alive': Count: 1 Sum: 1.000
    46  [2015-09-17 16:59:40 -0700 PDT][C] 'nomad.serf.member.join': Count: 1 Sum: 1.000
    47  [2015-09-17 16:59:40 -0700 PDT][C] 'nomad.raft.barrier': Count: 1 Sum: 1.000
    48  [2015-09-17 16:59:40 -0700 PDT][C] 'nomad.raft.apply': Count: 1 Sum: 1.000
    49  [2015-09-17 16:59:40 -0700 PDT][C] 'nomad.nomad.rpc.query': Count: 2 Sum: 2.000
    50  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.serf.queue.Query': Count: 6 Sum: 0.000
    51  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.fsm.register_node': Count: 1 Sum: 1.296
    52  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.serf.queue.Intent': Count: 6 Sum: 0.000
    53  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.runtime.gc_pause_ns': Count: 8 Min: 126492.000 Mean: 821516.875 Max: 3126670.000 Stddev: 1139250.294 Sum: 6572135.000
    54  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.raft.leader.dispatchLog': Count: 3 Min: 0.007 Mean: 0.018 Max: 0.039 Stddev: 0.018 Sum: 0.054
    55  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.leader.reconcileMember': Count: 1 Sum: 0.007
    56  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.leader.reconcile': Count: 1 Sum: 0.025
    57  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.raft.fsm.apply': Count: 1 Sum: 1.306
    58  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.client.get_allocs': Count: 1 Sum: 0.110
    59  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.worker.dequeue_eval': Count: 29 Min: 0.003 Mean: 363.426 Max: 503.377 Stddev: 228.126 Sum: 10539.354
    60  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.serf.queue.Event': Count: 6 Sum: 0.000
    61  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.raft.commitTime': Count: 3 Min: 0.013 Mean: 0.037 Max: 0.079 Stddev: 0.037 Sum: 0.110
    62  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.leader.barrier': Count: 1 Sum: 0.071
    63  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.client.register': Count: 1 Sum: 1.626
    64  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.nomad.eval.dequeue': Count: 21 Min: 500.610 Mean: 501.753 Max: 503.361 Stddev: 1.030 Sum: 10536.813
    65  [2015-09-17 16:59:40 -0700 PDT][S] 'nomad.memberlist.gossip': Count: 12 Min: 0.009 Mean: 0.017 Max: 0.025 Stddev: 0.005 Sum: 0.204
    66  ```
    67  
    68  # Key Metrics
    69  
    70  When telemetry is being streamed to statsite or statsd, `interval` is defined to
    71  be their flush interval. Otherwise, the interval can be assumed to be 10 seconds
    72  when retrieving metrics using the above described signals.
    73  
    74  <table class="table table-bordered table-striped">
    75    <tr>
    76      <th>Metric</th>
    77      <th>Description</th>
    78      <th>Unit</th>
    79      <th>Type</th>
    80    </tr>
    81    <tr>
    82      <td>`nomad.runtime.num_goroutines`</td>
    83      <td>Number of goroutines and general load pressure indicator</td>
    84      <td># of goroutines</td>
    85      <td>Gauge</td>
    86    </tr>
    87    <tr>
    88      <td>`nomad.runtime.alloc_bytes`</td>
    89      <td>Memory utilization</td>
    90      <td># of bytes</td>
    91      <td>Gauge</td>
    92    </tr>
    93    <tr>
    94      <td>`nomad.runtime.heap_objects`</td>
    95      <td>Number of objects on the heap. General memory pressure indicator</td>
    96      <td># of heap objects</td>
    97      <td>Gauge</td>
    98    </tr>
    99    <tr>
   100      <td>`nomad.raft.apply`</td>
   101      <td>Number of Raft transactions</td>
   102      <td>Raft transactions / `interval`</td>
   103      <td>Counter</td>
   104    </tr>
   105    <tr>
   106      <td>`nomad.raft.replication.appendEntries`</td>
   107      <td>Raft transaction commit time</td>
   108      <td>ms / Raft Log Append</td>
   109      <td>Timer</td>
   110    </tr>
   111    <tr>
   112      <td>`nomad.raft.leader.lastContact`</td>
   113      <td>Time since last contact to leader. General indicator of Raft latency</td>
   114      <td>ms / Leader Contact</td>
   115      <td>Timer</td>
   116    </tr>
   117    <tr>
   118      <td>`nomad.broker.total_ready`</td>
   119      <td>Number of evaluations ready to be processed</td>
   120      <td># of evaluations</td>
   121      <td>Gauge</td>
   122    </tr>
   123    <tr>
   124      <td>`nomad.broker.total_unacked`</td>
   125      <td>Evaluations dispatched for processing but incomplete</td>
   126      <td># of evaluations</td>
   127      <td>Gauge</td>
   128    </tr>
   129    <tr>
   130      <td>`nomad.broker.total_blocked`</td>
   131      <td>
   132          Evaluations that are blocked til an existing evaluation for the same job
   133          completes
   134      </td>
   135      <td># of evaluations</td>
   136      <td>Gauge</td>
   137    </tr>
   138    <tr>
   139      <td>`nomad.plan.queue_depth`</td>
   140      <td>Number of scheduler Plans waiting to be evaluated</td>
   141      <td># of plans</td>
   142      <td>Gauge</td>
   143    </tr>
   144    <tr>
   145      <td>`nomad.plan.submit`</td>
   146      <td>
   147          Time to submit a scheduler Plan. Higher values cause lower scheduling
   148          throughput
   149      </td>
   150      <td>ms / Plan Submit</td>
   151      <td>Timer</td>
   152    </tr>
   153    <tr>
   154      <td>`nomad.plan.evaluate`</td>
   155      <td>
   156          Time to validate a scheduler Plan. Higher values cause lower scheduling
   157          throughput. Similar to `nomad.plan.submit` but does not include RPC time
   158          or time in the Plan Queue
   159      </td>
   160      <td>ms / Plan Evaluation</td>
   161      <td>Timer</td>
   162    </tr>
   163    <tr>
   164      <td>`nomad.worker.invoke_scheduler.<type>`</td>
   165      <td>Time to run the scheduler of the given type</td>
   166      <td>ms / Scheduler Run</td>
   167      <td>Timer</td>
   168    </tr>
   169    <tr>
   170      <td>`nomad.worker.wait_for_index`</td>
   171      <td>
   172          Time waiting for Raft log replication from leader. High delays result in
   173          lower scheduling throughput
   174      </td>
   175      <td>ms / Raft Index Wait</td>
   176      <td>Timer</td>
   177    </tr>
   178    <tr>
   179      <td>`nomad.heartbeat.active`</td>
   180      <td>
   181          Number of active heartbeat timers. Each timer represents a Nomad Client
   182          connection
   183      </td>
   184      <td># of heartbeat timers</td>
   185      <td>Gauge</td>
   186    </tr>
   187    <tr>
   188      <td>`nomad.heartbeat.invalidate`</td>
   189      <td>
   190          The length of time it takes to invalidate a Nomad Client due to failed
   191          heartbeats
   192      </td>
   193      <td>ms / Heartbeat Invalidation</td>
   194      <td>Timer</td>
   195    </tr>
   196    <tr>
   197      <td>`nomad.rpc.query`</td>
   198      <td>Number of RPC queries</td>
   199      <td>RPC Queries / `interval`</td>
   200      <td>Counter</td>
   201    </tr>
   202    <tr>
   203      <td>`nomad.rpc.request`</td>
   204      <td>Number of RPC requests being handled</td>
   205      <td>RPC Requests / `interval`</td>
   206      <td>Counter</td>
   207    </tr>
   208    <tr>
   209      <td>`nomad.rpc.request_error`</td>
   210      <td>Number of RPC requests being handled that result in an error</td>
   211      <td>RPC Errors / `interval`</td>
   212      <td>Counter</td>
   213    </tr>
   214  </table>
   215  
   216  # Client Metrics
   217  
   218  The Nomad client emits metrics related to the resource usage of the allocations
   219  and tasks running on it and the node itself.  Operators have to explicity turn
   220  on publishing host and allocation metrics. Publishing allocation and host
   221  metrics can be turned on by setting the value of `publish_allocation_metrics`
   222  `publish_node_metrics` to `true`.
   223  
   224  
   225  By default the collection interval is 1 second but it can be changed by the
   226  changing the value of the `collection_interval` key in the `telemetry`
   227  configuration block.  
   228  
   229  Please see the [agent configuration](/docs/agent/config.html#telemetry_config)
   230  page for more details.
   231  
   232  ## Host Metrics
   233  
   234  <table class="table table-bordered table-striped">
   235    <tr>
   236      <th>Metric</th>
   237      <th>Description</th>
   238      <th>Unit</th>
   239      <th>Type</th>
   240    </tr>
   241    <tr>
   242      <td>`nomad.client.host.memmory.<HostID>.total`</td>
   243      <td>Total amount of physical memory on the node</td>
   244      <td>Bytes</td>
   245      <td>Gauge</td>
   246    </tr>
   247    <tr>
   248      <td>`nomad.client.host.memmory.<HostID>.available`</td>
   249      <td>Total amount of memory available to processes which includes free and
   250      cached memory</td>
   251      <td>Bytes</td>
   252      <td>Gauge</td>
   253    </tr>
   254    <tr>
   255      <td>`nomad.client.host.memory.<HostID>.used`</td>
   256      <td>Amount of memory used by processes</td>
   257      <td>Bytes</td>
   258      <td>Gauge</td>
   259    </tr>
   260    <tr>
   261      <td>`nomad.client.host.memory.<HostID>.free`</td>
   262      <td>Amount of memory which is free</td>
   263      <td>Bytes</td>
   264      <td>Gauge</td>
   265    </tr>
   266    <tr>
   267      <td>`nomad.client.uptime.<HostID>`</td>
   268      <td>Uptime of the host running the Nomad client</td>
   269      <td>Seconds</td>
   270      <td>Gauge</td>
   271    </tr>
   272    <tr>
   273      <td>`nomad.client.host.cpu.<HostID>.<CPU-Core>.total`</td>
   274      <td>Total CPU utilization</td>
   275      <td>Percentage</td>
   276      <td>Gauge</td>
   277    </tr>
   278    <tr>
   279      <td>`nomad.client.host.cpu.<HostID>.<CPU-Core>.user`</td>
   280      <td>CPU utilization in the user space</td>
   281      <td>Percentage</td>
   282      <td>Gauge</td>
   283    </tr>
   284    <tr>
   285      <td>`nomad.client.host.cpu.<HostID>.<CPU-Core>.system`</td>
   286      <td>CPU utilization in the system space</td>
   287      <td>Percentage</td>
   288      <td>Gauge</td>
   289    </tr>
   290    <tr>
   291      <td>`nomad.client.host.cpu.<HostID>.<CPU-Core>.idle`</td>
   292      <td>Idle time spent by the CPU</td>
   293      <td>Percentage</td>
   294      <td>Gauge</td>
   295    </tr>
   296    <tr>
   297      <td>`nomad.client.host.disk.<HostID>.<Device-Name>.size`</td>
   298      <td>Total size of the device</td>
   299      <td>Bytes</td>
   300      <td>Gauge</td>
   301    </tr>
   302    <tr>
   303      <td>`nomad.client.host.disk.<HostID>.<Device-Name>.used`</td>
   304      <td>Amount of space which has been used</td>
   305      <td>Bytes</td>
   306      <td>Gauge</td>
   307    </tr>
   308    <tr>
   309      <td>`nomad.client.host.disk.<HostID>.<Device-Name>.available`</td>
   310      <td>Amount of space which is available</td>
   311      <td>Bytes</td>
   312      <td>Gauge</td>
   313    </tr>
   314    <tr>
   315      <td>`nomad.client.host.disk.<HostID>.<Device-Name>.used_percent`</td>
   316      <td>Percentage of disk space used</td>
   317      <td>Percentage</td>
   318      <td>Gauge</td>
   319    </tr>
   320    <tr>
   321      <td>`nomad.client.host.disk.<HostID>.<Device-Name>.inodes_percent`</td>
   322      <td>Disk space consumed by the inodes</td>
   323      <td>Percent</td>
   324      <td>Gauge</td>
   325    </tr>
   326  </table>
   327  
   328  ## Allocation Metrics 
   329  
   330  <table class="table table-bordered table-striped">
   331    <tr>
   332      <th>Metric</th>
   333      <th>Description</th>
   334      <th>Unit</th>
   335      <th>Type</th>
   336    </tr>
   337    <tr>
   338      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.memory.rss`</td>
   339      <td>Amount of RSS memory consumed by the task</td>
   340      <td>Bytes</td>
   341      <td>Gauge</td>
   342    </tr>
   343    <tr>
   344      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.memory.cache`</td>
   345      <td>Amount of memory cached by the task</td>
   346      <td>Bytes</td>
   347      <td>Gauge</td>
   348    </tr>
   349    <tr>
   350      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.memory.swap`</td>
   351      <td>Amount of memory swapped by the task</td>
   352      <td>Bytes</td>
   353      <td>Gauge</td>
   354    </tr>
   355    <tr>
   356      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.memory.max_usage`</td>
   357      <td>Maximum amount of memory ever used by the task</td>
   358      <td>Bytes</td>
   359      <td>Gauge</td>
   360    </tr>
   361    <tr>
   362      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.memory.kernel_usage`</td>
   363      <td>Amount of memory used by the kernel for this task</td>
   364      <td>Bytes</td>
   365      <td>Gauge</td>
   366    </tr>
   367    <tr>
   368      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.memory.kernel_max_usage`</td>
   369      <td>Maximum amount of memory ever used by the kernel for this task</td>
   370      <td>Bytes</td>
   371      <td>Gauge</td>
   372    </tr>
   373    <tr>
   374      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.cpu.total_percent`</td>
   375      <td>Total CPU resources consumed by the task across all cores</td>
   376      <td>Percentage</td>
   377      <td>Gauge</td>
   378    </tr>
   379    <tr>
   380      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.cpu.system`</td>
   381      <td>Total CPU resources consumed by the task in the system space</td>
   382      <td>Percentage</td>
   383      <td>Gauge</td>
   384    </tr>
   385    <tr>
   386      <td>`nomad.client.allocs.<Job>.TaskGroup>.<AllocID>.<Task>.cpu.user`</td>
   387      <td>Total CPU resources consumed by the task in the user space</td>
   388      <td>Percentage</td>
   389      <td>Gauge</td>
   390    </tr>
   391    <tr>
   392      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.cpu.throttled_time`</td>
   393      <td>Total time that the task was throttled</td>
   394      <td>Nanoseconds</td>
   395      <td>Gauge</td>
   396    </tr>
   397    <tr>
   398      <td>`nomad.client.allocs.<Job>.<TaskGroup>.<AllocID>.<Task>.cpu.total_ticks`</td>
   399      <td>CPU ticks consumed by the process in the last collection interval</td>
   400      <td>Integer</td>
   401      <td>Gauge</td>
   402    </tr>
   403  </table>
   404  
   405  # Metric Types
   406  
   407  <table class="table table-bordered table-striped">
   408    <tr>
   409      <th>Type</th>
   410      <th>Description</th>
   411      <th>Quantiles</th>
   412    </tr>
   413    <tr>
   414      <td>Gauge</td>
   415      <td>
   416          Gauge types report an absolute number at the end of the aggregation
   417          interval
   418      </td>
   419      <td>false</td>
   420    </tr>
   421    <tr>
   422      <td>Counter</td>
   423      <td>
   424          Counts are incremented and flushed at the end of the aggregation
   425          interval and then are reset to zero
   426      </td>
   427      <td>true</td>
   428    </tr>
   429    <tr>
   430      <td>Timer</td>
   431      <td>
   432          Timers measure the time to complete a task and will include quantiles,
   433          means, standard deviation, etc per interval.
   434      </td>
   435      <td>true</td>
   436    </tr>
   437  </table>