Round Robin Database
The software writes temperature data into a Round Robin Database (RRD) stored in
A round robin database is a fixed size database meaning you can leave the data logger on for years and you’ll never run out of disk (SD card) space. By knowing what you want to monitor ahead of time (your job to configure) and what time series you want to aggregate over (my job to program), we can initialise the database and its size will remain the same.
Because we have to define what’s in the RRD upfront, you’ll have to delete the
~/.temperature/temperature.rdd if the configuration changes. Starting the server for the first time will initialise the RRD; it will create “archives” for every machine it’s expecting to monitor. These are found in
So, if you need to add additional machines after running the server for the first time, you’ll need to do the following.
A change in the configuration after the RRD has been created will not recreate or delete it - the configuration will be ignored until the RRD is deleted and the software restarted.