Best practices regarding deadman checks


I’m looking for some architectural guidance regarding checks. I’ve got a large fleet – 10,000s of devices – that report heartbeat data every 2.5 minutes. I’d like to use a very simple deadman check to listen and report on the health of those devices.

Is there an upper limit on the number of checks given Influx running on AWS with the default configuration? Could I set up a check for each device and feel confident that InfluxDB will handle it?


Hello @david.garrett,
Welcome! Are you running v2 with AWS timestream?
Please note v2 is no longer offered. The url says cloud2 but really it’s v3 unless youre using timestream with influxdb. If this is the case, flux and the alert system is no longer available.
You’d have to use another tool.
I think Grafana can confidently handle that type of alerting though.

Or are you a paying customer?
can you please check your version on the bottom right of the home page.

Yes, I spun up an instance of v2 with AWS timestream, but if that’s at EOL/EOS, then I would consider running something else.