Engagement Model · Day-2 Operations · SLAs

How Managed
OpenStack Works

A structured engagement from initial assessment through ongoing operations. We deploy production-grade OpenStack on your hardware and operate it continuously — upgrades, monitoring, patching, and capacity planning included.

The Process

Five Phases to Production

From first conversation to a fully managed private cloud — typically 2 to 4 weeks.

1

Assessment

We review your existing hardware, network topology, workload requirements, and compliance constraints. You get a clear architecture proposal and a deployment timeline — no guesswork.

2

Design

Control plane layout, storage topology, network architecture (OVN/Cilium), HA strategy, and monitoring stack. Documented and reviewed with your team before a single package is installed.

3

Deployment

Automated, repeatable deployment of the full OpenStack stack: Nova, Neutron, Cinder, Keystone, Ceph, Prometheus, Grafana. HA control plane with no single point of failure.

4

Handover

Your team gets API credentials, dashboard access, operational runbooks, and a walkthrough of the deployed architecture. You start provisioning workloads immediately.

5

Ongoing Operations

This is where the real value begins. We monitor, patch, upgrade, and capacity-plan your cloud 24/7. You consume the API — we keep the platform healthy.

Responsibility Matrix

What We Manage vs. What You Manage

Clear boundaries. No ambiguity about who owns what.

SDcloud Manages

  • OpenStack control plane (all core services)
  • Ceph storage cluster health and rebalancing
  • OpenStack upgrades between releases
  • Security patching (OS and OpenStack)
  • 24/7 monitoring, alerting, and incident response
  • Capacity planning and scale-out recommendations
  • Network infrastructure (OVN/Cilium, load balancers)
  • Backup of control plane state and configuration
  • Certificate management and rotation

You Manage

  • Virtual machines and workloads running on the platform
  • Applications deployed inside your instances
  • Guest OS patching within your VMs
  • Application-level monitoring and logging
  • User and project management (Keystone tenants)
  • Data and application-level backups
  • Hardware procurement and data center facility
Upgrades

Rolling Upgrades Between OpenStack Releases

OpenStack releases every six months. Most self-managed deployments fall behind within a year and never catch up. Managed OpenStack keeps your cloud current — every release, on schedule.

Our upgrade process is tested in staging environments before touching production. Control plane services are upgraded one at a time with rolling restarts. Compute nodes are drained and upgraded without VM downtime. The entire process is automated, auditable, and reversible.

Upgrade Process

  • Pre-upgrade validation in a staging environment
  • Rolling control plane upgrades (no API downtime)
  • Compute node drain and upgrade (live migration of VMs)
  • Ceph version alignment and OSD upgrades
  • Post-upgrade validation and smoke testing
  • Rollback plan tested before every upgrade begins

What Gets Upgraded

OpenStack Services

Nova · Neutron · Cinder · Keystone · Glance · Horizon · Octavia · Heat

Storage

Ceph (Reef, Squid, and beyond) · OSD · RGW · MDS

Networking

OVN · Cilium · HAProxy · Keepalived

Observability

Prometheus · Grafana · Alertmanager · OpenSearch

Why this matters: Running an unsupported OpenStack release means no security patches, no bug fixes, and a growing gap between your deployment and upstream. Every release you skip makes the next upgrade exponentially harder. Staying current is not optional — it is the single most important operational practice.

Service Levels

SLA Tiers

Choose the level of support that matches your operational requirements.

StandardPremiumMission Critical
Monitoring24/7 automated24/7 automated24/7 automated + dedicated
Incident responseBusiness hours24/724/7 (15-min SLA)
Control plane uptime99.5%99.9%99.95%
Upgrade frequencyEvery releaseEvery releaseEvery release
Capacity reviewsQuarterlyMonthlyContinuous
Dedicated engineerNoNoYes
Architecture reviewsAnnualSemi-annualQuarterly
Day-2 Operations

What's Included After Go-Live

Deployment is just the beginning. The real work is keeping it running.

Monitoring & Alerting

Prometheus-based monitoring of every OpenStack service, Ceph OSD, and network component. Alertmanager routes critical alerts to our on-call engineers. You get read access to all dashboards.

Patching & Security

OS-level and OpenStack security patches applied on a regular cadence. Critical CVEs addressed within 24 hours. All changes are tested, staged, and applied with zero or minimal downtime.

Capacity Planning

Proactive monitoring of compute, storage, and network utilization trends. We flag capacity constraints before they become incidents and provide scale-out recommendations with lead time.

Incident Response

When something breaks, our engineers respond — not a ticket queue. Root cause analysis provided for every significant incident. Post-incident reports within 48 hours.

Configuration Management

All infrastructure configuration is version-controlled and reproducible. Changes go through a review process. Drift detection ensures what's running matches what's declared.

Backup & Recovery

Control plane state, Ceph configuration, and critical service data are backed up continuously. Recovery procedures are documented and tested. RTO and RPO targets defined per SLA tier.

Get Started

Ready to Hand Off the Operational Burden?

Tell us about your hardware and workload requirements. We'll come back with an architecture proposal and a deployment timeline.