As we have Prometheus 1.6 built in now we might have incorrect information in FAQ or might not follow the best practices as the release notes state:
[FEATURE] New flag -storage.local.target-heap-size to tell Prometheus about the desired heap size. This deprecates the flags -storage.local.memory-chunks and -storage.local.max-chunks-to-persist, which are kept for backward compatibility.
If I understand correctly this makes the limit set up of RAM usage in 1.6, how does that affect the FAQ?
The limit affects only memory reserved for data chunks. Actual RAM usage by Prometheus is higher. It is recommended to have at least three times more memory than the expected memory taken up by data chunks.
[image: Inline image 1]
- Update supervisord.conf for Prometheus 1.6 syntax
- Update documentation with more appropriate wording modeled on guidance from prometheus.io
set -storage.local.target-heap-size to a value of two thirds of the physical memory limit Prometheus should not exceed