Server Monitoring

InfluxDB can display statistical and diagnostic information about each node. This information can be very useful for troubleshooting and performance monitoring.


To see node stats execute the command SHOW STATS.

The statistics returned by SHOW STATS are stored in memory only, and are reset to zero when the node is restarted.


To see node diagnostics execute the command SHOW DIAGNOSTICS. This returns information such as build information, uptime, hostname, server configuration, memory usage, and Go runtime diagnostics.

Internal Monitoring

InfluxDB also writes statistical and diagnostic information to database named _internal, which records metrics on the internal runtime and service performance. The _internal database can be queried and manipulated like any other InfluxDB database. Check out the monitor service README and the internal monitoring blog post for more detail.

Useful Commands

Below are a collection of commands to find useful performance metrics about your InfluxDB instance.

To find the number of points per second being written to the instance. Must have the monitor service enabled:

$ influx -execute 'select derivative(pointReq, 1s) from "write" where time > now() - 5m' -database '_internal' -precision 'rfc3339'

To find the number of writes separated by database since the beginnning of the log file:

grep 'POST' /var/log/influxdb/influxd.log | awk '{ print $10 }' | sort | uniq -c

Or, for systemd systems logging to journald:

journalctl -u influxdb.service | awk '/POST/ { print $10 }' | sort | uniq -c

© 2015 InfluxData, Inc.
Licensed under the MIT license.