InvalidArgument: Points must be written in order.

[vsekhar]

After starting two local instances of my server and letting them run for about a minute, I get the following error about every minute:

2020/12/31 21:41:40 rpc error: code = InvalidArgument desc = One or more TimeSeries could not be written: Points must be written in order. One or more of the points specified had an older start time than the most recent point.: timeSeries[1-12]

Based on some earlier comments, I tried ensuring I am setting the host and service names without success.

r.peerCount = metric.Must(meter).NewInt64ValueRecorder(
		"peers",
		metric.WithDescription("Number of peers found and joined"),
	).Bind([]label.KeyValue{
		label.String("host.id", nodeName),
		label.String("service.name", "peer_service"),
		label.String("service.instance.id", nodeName),
	}...)

I also tried setting the OC_RESOURCE_LABEL=host.id=instance_123 environment variable without success.

What is the correct way to provide an instance_id when not running on GCE or GKE? Or is there some other issue here?

[punya]

Is nodeName the same between the two local instances of your server?

[vsekhar]

It is different. Each instance randomly generates a name suffix.

[dashpole]

Sorry for the slow response. Which monitored resource is your metric being written to?

[dashpole]

I'm going to close this for now, but feel free to reopen if you are still having problems. For others that find this, I would recommend using the gcp resource detector: https://github.com/open-telemetry/opentelemetry-go-contrib/tree/main/detectors/gcp. Also, make sure the export interval of your periodic reader is at least 10 seconds.