Services: 'mysql', 'rabbitmq', 'haproxy', 'memcached', 'apache'
Source: 'heartbeat'
The AFD emits an afd_service_metric every 30 seconds, sets the status to UNKNOWN
for services without metric in this interval otherwise the status is OKAY.
Note about why not set to DOWN instead of UNKNOWN: the reason is that
collectd itself could be DOWN but not the monitored service.
UNKNOWN status message:
:Type: heka.sandbox.afd_service_metric
:Logger: service_heartbeat_filter
:Payload: {"alarms":[{"periods":0,"tags":{},"severity":"UNKNOWN","window":"30","operator":"==","function":"last","fields":[{"name":"service","value":"rabbitmq"}],"metric":"rabbitmq*","message":"No metric received from the service","threshold":0,"value":0}]}
:Fields:
| name:"source" type:string value:"heartbeat"
| name:"service" type:string value:"rabbitmq"
| name:"tag_fields" type:string value:"service"
| name:"name" type:string value:"service_status"
| name:"interval" type:string value:"30"
| name:"value" type:double value:2
OKAY status message:
:Payload: {"alarms":[]}
:Fields:
| name:"source" type:string value:"heartbeat"
| name:"service" type:string value:"rabbitmq"
| name:"tag_fields" type:string value:"service"
| name:"name" type:string value:"service_status"
| name:"interval" type:string value:"30"
| name:"value" type:double value:0
Change-Id: I42be4e3bf3d59d4cba6ebf838c8553a4f1b99087
202d4fc935