1-Minute Polling
LibreNMS can poll data at the interval that you select.
Read these conditions first:
- For faster up and down alerts, Fast Ping is a much easier method.
- If you change the interval from the default of 300 seconds, you must also change the cron entry for
poller-wrapper.py. - The polling MUST complete within the heartbeat step value. Open
/pollerin your web interface to see the current value. - The change applies only to RRD files that LibreNMS creates after the change.
- The change applies to all storage mechanisms, such as MySQL, RRD, and InfluxDB. A lower value increases the disk space for MySQL and InfluxDB.
- Configure some performance optimizations. Each device has many graphs, and each device writes all its graphs to the disk every minute. The RRDCached configuration is the most important one, because it saves many write IOPS.
To make the change, open /settings/poller/rrdtool/ in your web interface. Select RRDTool Setup. Then set the two values for the step interval and the heartbeat interval:
- Step is the interval between two data inserts. For 1-minute polling, set this value to 60.
- Heartbeat is the time to wait for data before LibreNMS records a null value. An example value is 120 seconds.
Converting existing RRD files
We supply a basic script. This script converts the default RRD files to your step value and your heartbeat value. Back up your RRD files before you run the script. The script runs for one device or for all devices.
The RRD files must be available on the server that runs this script.
lnms maintenance:rrd-step
This command shows the help information. To run the script for localhost, run:
lnms maintenance:rrd-step localhost