Managed Prometheus Hosting — production-ready from $15 a month
Monitoring system with time-series database. Deployed on your own dedicated instance in AWS, Azure, or GCP, kept patched, backed up, and monitored by ManageStacks — standard Prometheus, no lock-in.
Prometheus on ManageStacks is the de-facto open-source metrics platform (a CNCF graduated project) deployed to your own AWS, Azure, or GCP region — priced flat at $29 per instance per month regardless of metric volume, with Alertmanager included, service discovery for Kubernetes/EC2/Consul, and long-term storage via Thanos or Mimir on Business+ plans. Materially cheaper than Datadog or Grafana Cloud metrics for infra observability, and your metric data stays in your cloud.
PromQL · Pull-based scraping
What does Prometheus do, and why do teams deploy it?
Prometheus is the de-facto open-source metrics collection and alerting platform — a CNCF graduated project used by Kubernetes, most SRE teams, and the majority of production Docker/container deployments. It scrapes metrics over HTTP from configured targets, stores them in a purpose-built time-series database, and evaluates alerting rules on a schedule.
Prometheus's multi-dimensional data model (metrics identified by name + key/value labels), its expressive PromQL query language, and its pull-based scrape model make it the standard interface for infrastructure and application metrics. Every major cloud vendor, Kubernetes operator, and observability tool speaks Prometheus.
Running production Prometheus yourself means running the prometheus binary, sizing local TSDB storage for your retention window (2 weeks is the default; longer retention means bigger disks), configuring service discovery for your dynamic environment (K8s, EC2, Consul), running Alertmanager for notification routing, and — for retention beyond a few weeks or high availability — running Thanos or Grafana Mimir on top for long-term storage. ManageStacks handles all of that.
- Multi-dimensional data model with labels + PromQL query language
- Pull-based scraping over HTTP with configurable intervals
- Alertmanager included — routing, silencing, and grouping
- Service discovery: Kubernetes, EC2, GCE, Azure, Consul, DNS, file-based
- Recording rules for pre-computed queries used in dashboards
- Efficient TSDB storage with compaction and downsampling
Monitoring system with time-series database
What does managed Prometheus hosting cost?
Flat per-app pricing, in your chosen AWS, Azure, or GCP region. No per-user pricing — a busy deployment costs the same as a quiet one.
Starter
Staging and internal tools. Dedicated instance, TLS, daily backups, managed upgrades.
Standard
Production workloads. Adds monitoring, staging environment, region choice, priority support.
Business
High-traffic and compliance workloads. Adds a high-availability replica and same-day support.
24×7 SRE retainer
Round-the-clock on-call across every hosted application, for teams that need a pager answered at 3am.
Self-hosting Prometheus vs managed — what does it really cost?
The software is free. The engineer-hours are not.
Running it yourself
- Install Prometheus + Alertmanager on a VM; size local TSDB storage manually
- Configure service discovery for your dynamic environment by hand
- Set up Grafana separately and point it at Prometheus
- Add Thanos or Mimir + object storage for long-term retention
- Test upgrades manually; build federation topology for multi-cluster
On ManageStacks
- Subscribe through your AWS, Azure, or GCP marketplace
- Prometheus comes up with Alertmanager, TSDB, service discovery, and Grafana wired in
- Thanos/Mimir available on Business+ for object-storage-backed long-term retention
- Pre-built Grafana dashboards for node, postgres, mysql, redis, kubernetes exporters
- Version upgrades handled by us with zero-downtime rolling restarts on HA plans
Prometheus on ManageStacks vs the alternatives
How Prometheus on ManageStacks compares to the vendor-hosted OSS observability offering and the two dominant APM SaaS platforms.
| Prometheus on ManageStacks | Grafana Cloud (metrics) | Datadog | New Relic | |
|---|---|---|---|---|
| Deployment | Managed on your AWS, Azure, or GCP | Vendor-hosted | Vendor-hosted | Vendor-hosted |
| Data residency | Your cloud region | Vendor infrastructure | Vendor infrastructure | Vendor infrastructure |
| Pricing basis | Flat per instance | Per active series + ingest | Per host + per custom metric | Per user + per GB ingested |
| Long-term retention | Thanos/Mimir on Business+ | Included (Mimir-backed) | Included (paid tiers) | Included |
| Open source | Yes (Apache 2.0, CNCF) | Yes (Apache, hosted) | No (proprietary) | No (proprietary) |
| Alerting | Alertmanager included | Included | Included | Included |
Provisioning, upgrades, backups and monitoring on your team’s plate.
What does running Prometheus yourself involve?
ManageStacks deploys Prometheus with persistent TSDB storage, Alertmanager, service discovery (Kubernetes/EC2/Consul), and Grafana with pre-built dashboards. On Business+ we add Thanos or Mimir for long-term retention. We handle version upgrades, storage scaling, and the federated topology for multi-cluster observability.
Prometheus key numbers
How long from subscribing to a live instance?
Subscribe
Subscribe to ManageStacks through your AWS, Azure, or GCP marketplace.
Provision
Prometheus + Alertmanager + Grafana spin up together with pre-built dashboards — typically 3-5 minutes.
Point scrape targets
Add scrape configs for your services and infrastructure (K8s service discovery, EC2 tags, static targets). Point exporters at Prometheus.
Alert + visualise
Configure Alertmanager routing (Slack, PagerDuty, email), build dashboards in Grafana, and set up recording rules for expensive queries.
When is self-hosting Prometheus the right answer instead?
“Managed hosting is not always the correct call.”
Self-host when a platform team already runs the infrastructure and on-call rotation to operate Prometheus at genuinely low marginal cost. Self-host when compliance requires an air-gapped or on-premises deployment that no hosted option can satisfy. And self-host when the deployment depends on heavy customisation with a fast internal build-deploy loop, because an internal release process will beat any managed change process.
For everyone else — teams whose engineers have better things to do than shepherd upgrades — managed hosting is cheaper than the hours it replaces.
Which cloud should Prometheus run on — AWS, Azure or GCP?
For most workloads, the choice of cloud matters less than proximity: run Prometheus in the same cloud and region as the applications and data it talks to, because every request between them adds a round trip. The underlying compute performs equivalently across AWS, Azure, and GCP.
In practice, an existing cloud footprint decides it. All plans support all three clouds, and moving regions later is a scheduled migration, not a rebuild.
Deepest managed-service catalog, default when there's no existing footprint
Best fit for teams already on Microsoft 365 or Entra ID
Strongest for data/analytics-adjacent workloads
Every plan supports AWS, Azure, and GCP — region choice included.
Common questions about Prometheus on ManageStacks
How long does it take to deploy Prometheus on ManageStacks?
Under 5 minutes. ManageStacks provisions Prometheus with Alertmanager, persistent TSDB storage, service-discovery configuration, and Grafana with pre-built dashboards for the standard exporters (node, postgres, mysql, kubernetes).
How does this compare to Grafana Cloud metrics or Datadog?
Grafana Cloud metrics is priced per active series (~$0.30 per 1,000 series/month at scale) + per ingest. Datadog is per host + per custom metric. Both grow unpredictably. ManageStacks Prometheus is flat $29 per instance regardless of series count. For infrastructure and application metrics under 1 million active series, self-hosted on ManageStacks is materially cheaper. Grafana Cloud / Datadog wins when you want true elastic scale, cross-tenant analytics, or vendor-managed SLOs.
How is long-term metric retention handled?
Standard Prometheus retention is limited by local disk. On Business+ plans we deploy Thanos or Grafana Mimir alongside for object-storage-backed long-term retention — years of metric history queryable from Grafana without changing the query interface. Downsampling reduces storage cost for older data.
Can I configure custom scrape targets and alerting rules?
Yes. Scrape configs and alerting rules manage through a Git-synced config repo (recommended) or directly through the dashboard. Recording rules for pre-computed dashboard queries, silences, and inhibition rules are all supported.
Does ManageStacks handle Prometheus and Alertmanager version upgrades?
Yes. We track upstream Prometheus and Alertmanager releases, test against a representative dataset, and roll forward with zero downtime on replicated setups.
Can I federate multiple Prometheus instances?
Yes. Federation (a top-level Prometheus scraping aggregate metrics from environment-specific Prometheuses) is a standard pattern for multi-cluster or multi-region setups. We configure the federation topology as part of onboarding.
What about high-availability Prometheus?
Standard plan is a single Prometheus with local TSDB. Business+ plans deploy 2+ Prometheus replicas scraping the same targets, with Thanos Query fronting them for deduplication — high-availability metrics with no gap on failover.
What if I want to migrate off?
Prometheus TSDB is portable — copy the data directory or use `promtool tsdb export`. If you're on Thanos or Mimir, the object storage bucket is portable to any Thanos/Mimir deployment. Migration off ManageStacks is a supported operation.
Run Prometheus without carrying the pager
Subscribe through your AWS, Azure, or GCP marketplace. We handle provisioning, SSL, monitoring, backups, updates, and security. From $15/mo.