Displays a heatmap for the specified nodes.

Syntax

CLI

maprcli node heatmap
    [ -cluster <cluster> ]
    [ -filter <filter> ]
    [ -view <view> ]
    -json | -long

REST

http[s]://<host>:<port>/rest/node/heatmap?<parameters> 

Parameters

Parameter

Description

cluster

The cluster on which to run the command.

filter

A filter specifying snapshots to preserve. See Filters for more information.

view

Name of the heatmap view to show:

  • status = Node status (the default view):
    • 0 = Healthy
    • 1 = Needs attention
    • 2 = Degraded
    • 3 = Maintenance
    • 4 = Critical
  • cpu = CPU utilization, as a percent from 0-100.
  • memory = Memory utilization, as a percent from 0-100. 
  • diskspace = MapR-FS disk space utilization, as a percent from 0-100. 
  • NODE_* = Status of various alarms: 0 if clear, 1 if raised. For example: NODE_ALARM_DISK_FAILURE or NODE_ALARM_SERVICE_CLDB_DOWN.
    You can return a complete list of supported alarm parameters by running:

    maprcli node heatmap -view

    See the lists of parameters below this table.

-json | -longThis command returns multiple levels of data. You must specify either JSON format or "long" format to see the full output.

Alarm Parameters

You can view the status of a number of different alarms, including the status of alarms for services down and alarms for other conditions on the cluster.

Service Down Alarms

NODE_ALARM_SERVICE_CLDB_DOWN
NODE_ALARM_SERVICE_FILESERVER_DOWN
NODE_ALARM_SERVICE_JT_DOWN
NODE_ALARM_SERVICE_TT_DOWN
NODE_ALARM_SERVICE_HBMASTER_DOWN
NODE_ALARM_SERVICE_HBREGION_DOWN
NODE_ALARM_SERVICE_WEBSERVER_DOWN
NODE_ALARM_SERVICE_NFS_DOWN
NODE_ALARM_SERVICE_HOSTSTATS_DOWN
NODE_ALARM_SERVICE_OOZIE_DOWN
NODE_ALARM_SERVICE_HUE_DOWN
NODE_ALARM_SERVICE_HTTPFS_DOWN
NODE_ALARM_SERVICE_BEESWAX_DOWN
NODE_ALARM_SERVICE_HIVEMETA_DOWN
NODE_ALARM_SERVICE_HS2_DOWN

Other Alarms

NODE_ALARM_DEBUG_LOGGING
NODE_ALARM_DISK_FAILURE
NODE_ALARM_VERSION_MISMATCH
NODE_ALARM_TIME_SKEW
NODE_ALARM_ROOT_PARTITION_FULL
NODE_ALARM_OPT_MAPR_FULL
NODE_ALARM_CORE_PRESENT
NODE_ALARM_HIGH_MFS_MEMORY
NODE_ALARM_PAM_MISCONFIGURED
NODE_ALARM_TT_LOCALDIR_FULL
NODE_ALARM_NO_HEARTBEAT
NODE_ALARM_MAPRUSER_MISMATCH
NODE_ALARM_DUPLICATE_HOSTID
NODE_ALARM_METRICS_WRITE_PROBLEM
NODE_ALARM_TOO_MANY_CONTAINERS

 

Output

In general, the heatmap output looks like this (in JSON format).

{
    status:"OK",
    data: [{
            "{{rackTopology}}" : {
                    "{{nodeName}}" : {{heatmapValue}},
                    "{{nodeName}}" : {{heatmapValue}},
                    "{{nodeName}}" : {{heatmapValue}},
                    ...
                },
            "{{rackTopology}}" : {
                    "{{nodeName}}" : {{heatmapValue}},
                    "{{nodeName}}" : {{heatmapValue}},
                    "{{nodeName}}" : {{heatmapValue}},
                    ...
                },
                ...
    }]
}

Output Fields

Field

Description

rackTopology

The topology for a particular rack.

nodeName

The name of the node in question.

heatmapValue

The value of the metric specified in the view parameter for this node, as an integer.

Examples

Display a heat map with the node status (default view) for the default rack:

maprcli node heatmap -json
{
	"timestamp":1422567293873,
    "timeofday":"2015-01-29 01:34:53.873 GMT-0800",
    "status":"OK",
    "total":1,
    "data":[
       {"/data/default-rack":{
        "centos24":2}
       }]
}

The equivalent REST API command would be:

https://r1n1.sj.us:8443/rest/node/heatmap

 

Display memory usage for the default rack:

maprcli node heatmap -view memory -json
{
    "timestamp":1422585976631,
    "timeofday":"2015-01-29 06:46:16.631 GMT-0800",
    "status":"OK",
    "total":1,
    "data":[
       {"/data/default-rack":{
        "centos24":71}
       }]
} 

The equivalent REST API command would be:

https://r1n1.sj.us:8443/rest/node/heatmap?view=memory

Display the value of NODE_ALARM_DISK_FAILURE for the default rack:

maprcli node heatmap -view NODE_ALARM_DISK_FAILURE -long
/data/default-rack  
{"centos24":0}