Period and Every in DeadMan Alert

InfluxDB 1 Kapacitor
1

We are using telegraf to monitor our servers and network devices in which telegraf sends data every 5 min, now we want to get a alert when any of our server did not send data for 15 min. below is the script i am using to achieve that.

var data = batch
|query(''' Select last(usage_user) from prod_hosts.autogen.cpu where cpu = 'cpu-total' ''')
    .period(15m)
    .every(5m)
    .groupBy(*)

data
|deadman(0.0, 15m)
    .stateChangesOnly()
    .id('{{ index .Tags "node" }}')
    .message('Server {{ .ID }} is OFFLINE')
    .messageField('message')
    .email()
    .to('XXXXXX')
    .log('/tmp/chronograf/deadman.log')

now when i use Kapacitor watch to check this task it is running fine and if i run the influxql query in db i get the result without the server for which i ave stopped telegraf service.

But It is not triggering any alerts, my initial though is it have something to do with time in deadman and period.

2

Hi @Amit_Kumar_Singh ,
in your script there will be an alert if the throughput drops below 0.0 points per 15m but that will never happen , i guess you will have to use deadman(1.0 , 15m ) for example ,

best regards ,