Prometheus 2.19.0 is now out, following on from 2.18.0 with many fixes and improvements.

The big new feature if that full chunks in the TSDB head block are now stored in mmaped, rather than on the heap. This reduces memory requirements particularly for higher frequency scraping, and allows for faster restarts. There's also been improvements in isolation performance.

Triton service discovery now supports global zones, the alert resend throttling is more tolerant of failures, and there's a new prometheus_remote_storage_remote_read_queries_total metric.

There’s also a number of improvements, bug fixes and other changes, see the release notes for more detail.


Wondering how to upgrade? Contact us.