See this version's CHANGELOG.md file for a full list of changes. From this release forward, we are attempting to track to versions of Lustre that significantly alter their procfs layout. This release has been tested against Lustre 2.10.1.

See this version's CHANGELOG.md file for a full list of changes. This release is to migrate a released version of the Lustre Exporter to dep for dependency tracking (from govendor).

See this version's CHANGELOG.md file for a full list of change. This is the last release that is intended to support IEEL 3.x versions. Going forward, master will track with upstream Lustre (starting with 2.10.1).

See this tag's CHANGELOG.md for a full report of the updates included in this version of the HPE Lustre Exporter. Note that there will still be some fine-tuning of labels prior to an official 1.0.0 release.

This is the initial pre-1.0 release of the Prometheus lustre_exporter.

We support metrics from MDT, OST, MDS, MGS, and OST-related LDLM procfs directories. If there are any metrics missing that you actively use, please file an issue and we'll address that if possible (or explain why we can't get that metric or chose not to).

The major concern right now would be environments with long retention windows and high job counts. Please don't hesitate to reach out if you run into a situation where Prometheus performance degrades below acceptable thresholds, with an explanation of the environment you're running in and an idea of how many jobs you're running and your configured retention timeframe in Prometheus.