Prometheus exporter for Lustre metrics.
go get github.com/HewlettPackard/lustre_exporter
cd $GOPATH/src/github.com/HewlettPackard/lustre_exporter
make
./lustre_exporter
TODO, but the current plan is to create flags that would define what node metrics are pulled from a running instance (e.g., OSS and MDS metrics would each have their own flag to disable). Also, we'd have flags to disable non-procfs metrics.
Design plans
- Export all proc data from all nodes running the Lustre Exporter that can function as a counter type (will save histogram-type work for later).
- STATUS: Mostly complete as far as we can tell. If anything you want is missing, please file an issue!
- Identify redundant data (if it exists).
- Deduplication would be done, at first, by enabling flags to identify the node type with a configuration flag.
- STATUS: No problems so far, saving for future work.
- Add in:
- Histogram data
- STATUS: We have some of this via the 'brw_stats' file. We have created the histograms within Grafana in early tests.
- Other data sources (CLI data that isn't present in /proc, for example). Users will be able to disable non-proc sources via a configuration flag.
- STATUS: Not yet started
To contribute to this HPE project, you'll need to fill out a CLA (Contributor License Agreement). If you would like to contribute anything more than a bug fix (feature, architectural change, etc), please file an issue and we'll get in touch with you to have you fill out the CLA.