Open xginn8 opened 4 years ago
Most of this work is done now. The recommendations we skipped were related to reducing memory/CPU footprint, which we'll deal with if it comes up.
We should prepare a set of defaults for low-memory/cpu situations, which may be required in specific usecases.
Another datapoint from the EtcherPro
netdata uses more cpu and emmc writes than collectd + influxdb + grafana;
Netdata config has some useful settings for running on embedded & IoT. We should review and use what makes sense: https://learn.netdata.cloud/docs/agent/performance#running-netdata-in-embedded-devices & https://learn.netdata.cloud/docs/agent/netdata-for-iot/