Knowing what your storage is doing, in the open

Storage that fails silently is worse than storage that fails loudly. 45Drives builds monitoring on the same open-source stack most of the infrastructure world already trusts: Prometheus for metrics collection, Grafana for dashboards and visualization, and Alertmanager for routing alerts to the people who need to see them.

Why this matters for a storage server

A proprietary monitoring dashboard locks your operational visibility to one vendor’s UI and one vendor’s alerting rules. Prometheus/Grafana/Alertmanager is the de facto standard observability stack across the rest of the infrastructure world — the same tools monitoring your storage servers likely already watch your applications, your network, and everything else in a modern operations team’s stack. One set of dashboards, one alerting pipeline, one skill set to maintain, instead of a separate silo just for storage.

How 45Drives uses it

45Drives publishes a set of open-source Ansible roles that deploy Prometheus, Alertmanager, and Grafana together, pre-wired to collect metrics from Ceph and ZFS storage nodes. Rather than building a proprietary dashboard from scratch, 45Drives automated the deployment of tools the broader industry already standardized on.

There’s also a lighter-weight cockpit-alerts module built directly into Houston UI for surfacing alerts inside the dashboard itself — worth noting this one doesn’t currently have a public source repository, unlike the Ansible roles and exporters below.

That’s backed by purpose-built Prometheus exporters 45Drives publishes for the parts of the stack that don’t have off-the-shelf metrics already: zfs_exporter, autotier_exporter for autotier tiering activity, radosgw_usage_exporter for Ceph RGW/S3 usage, and cephgeorep_exporter for CephFS remote-backup status.

Why that matters if you’re evaluating storage

• No dashboard lock-in. Metrics live in Prometheus’s standard format — queryable, exportable, and usable with any tool that speaks that ecosystem, not just 45Drives’ own UI.
• One observability stack for everything, not a separate proprietary tool just for storage — if your team already runs Prometheus/Grafana elsewhere, storage metrics slot right in.
• Alerting you control. Alertmanager routes to whatever you already use — email, Slack, PagerDuty, webhook — rather than being limited to whatever notification channels one vendor’s product happens to support.

Learn more

• prometheus.io — the official Prometheus project
• grafana.com — the official Grafana project
• 45Drives’ monitoring-stack repository

See Ceph and ZFS for the storage layers this monitoring stack is built to watch.