Service 02 · Operations

Operations & monitoring: watched around the clock, fully documented

In brief

At managed-cluster.net, 24/7 infrastructure monitoring covers six layers of your cluster: hardware, virtualization, operating system, services, application and network. Critical alerts reach on-call duty immediately, nights and public holidays included. You get a post-incident report for every incident and an operations report every month. Operations are included in the fixed monthly price; response targets go into the operations contract.

Monitoring layers

24/7 infrastructure monitoring across six layers

The managed monitoring service from managed-cluster.net covers every layer of your cluster: hardware, virtualization, operating system, services, application and network. Health checks from the user's perspective also test whether your application actually responds, beyond whether its processes are running. Operations & monitoring is one of four areas of our managed infrastructure operations, which together form a single operations contract.

Layers What we monitor and what happens when something fails
LayerWhat we watchWhen something fails
Hardware Cluster node servers, components, disks Defective components are replaced around the clock (hardware replacement via remote hands); disk replacements are documented in the post-incident report.
Virtualization Proxmox VE, quorum via Corosync, HA Manager If a node fails, the Proxmox HA Manager automatically restarts the affected virtual machines on another node after fencing, and on-call duty is alerted.
Operating system Nodes and the systems we run for you We fix the cause; security updates follow our managed patch management node by node, critical vulnerabilities first.
Services Platform services such as load balancers with health checking and Ceph storage The load balancer takes a failed node out of rotation; after a disk failure, Ceph restores the target redundancy on its own.
Application Health checks from the user's perspective: does your application respond? For shop clusters for Shopware and Magento, for example, we check the home page, search and checkout. The application itself remains your responsibility.
Network AS47263, redundant switches and uplinks, bonding/LACP on every node If one path fails, redundant network paths keep the cluster connected, and the fault triggers an alert (network monitoring on AS47263).

Alerting

24/7 alerting with an escalation chain and on-call IT operations

When 24/7 monitoring detects a fault in your cluster, it raises an alert automatically. The alert follows defined escalation paths and reaches on-call duty, at night, on weekends and on public holidays. The first signal comes from monitoring, not from your users. You also have a dedicated contact who knows your cluster.

Response to critical alerts

Critical alerts reach on-call duty immediately and are handled right away, around the clock. We set specific response and recovery targets together with you in the operations contract.

Diagram 01 From metric to post-incident report
  • Metrics, continuous
  • Alert and documentation
  • Layer monitored
  • Alert to on-call duty
Metrics from six layers converge in 24/7 monitoring, and an alert reaches on-call duty through the escalation chain. Remediation is followed by a post-incident report; every month brings an operations report.
Diagram as text
  1. Hardware, virtualization, operating system, services, application and network deliver metrics continuously, complemented by health checks from the user's perspective.
  2. Monitoring evaluates them around the clock; when a fault occurs, an alert reaches on-call duty through the escalation chain.
  3. We analyze and fix the fault, and you receive a post-incident report with cause and actions.
  4. Every month, the operations report summarizes availability, incidents and utilization; targets are set in the operations contract.

Incident management

Incident management in hosting: from alert to post-incident report

In our incident management, every incident goes through the same documented process: detection, alert, analysis, remediation and a written post-incident report covering the cause and the actions derived from it. That lets you trace what happened and what follows from it, whether a node failed or a network link was disrupted.

  1. DETECTION. A fault on one of the six layers or a failed health check.
  2. ALERT. Through the escalation chain to on-call duty, critical alerts immediately.
  3. ANALYSIS. We narrow down the cause: hardware, platform, network or application.
  4. REMEDIATION. We restore redundancy; defective hardware is replaced through remote hands.
  5. POST-INCIDENT REPORT. Cause and derived actions, in writing.

Priorities and target times are set in the operations contract. What the cluster handles on its own when a node fails is shown in the automatic failover sequence.

What the post-incident report records

The post-incident report sets out in writing what happened during an incident, what caused it and which actions we derive from it. You receive it after every incident for your own records.

Your organization's obligations, such as notifying authorities or business partners, stay with you. The post-incident report is written only after the incident and does not replace any such notification. What we take on in operations and what stays with you is set out in our security and compliance responsibility overview.

Service levels

Hosting SLA: what the operations contract specifies

In the operations contract, we agree with you on the availability target, response times by priority level, service hours with escalation paths, and the recovery targets RTO (recovery time objective) and RPO (recovery point objective). We do not publish blanket percentages, because what we can commit to depends on the architecture: node count, service redundancy and site design.

Operations contract What the SLA covers and what you receive in day-to-day operations
ItemWhat is agreedWhat you receive
Availability target Your platform's availability, derived from its architecture A monthly operations report including availability; SLA reports on top in the Business package
Response times Target times by priority level; critical alerts around the clock A post-incident report with cause and actions for every incident
Service hours and escalation paths When we can be reached and how an alert escalates A dedicated contact who knows your cluster
Recovery targets RTO and RPO for the agreed failure scenarios Failover tests with a written log and regular restore tests

Values per priority level are agreed per environment in the operations contract. As of

How much downtime an availability target allows per year is shown by our availability calculator; the terms are explained under RTO and RPO in detail. Tell us about your requirements; we will tell you plainly which of them we can already provide evidence for.

Reports

Monthly operations report, post-incident reports and live metrics

Every managed Proxmox cluster comes with a monthly operations report covering availability, incidents and utilization, plus our recommendations. The Business package adds SLA reports and extended monitoring with escalation. Every incident also gets its own post-incident report covering the cause and the actions derived from it. On request, we set up read access to a dashboard with live metrics.

  • Post-incident report: cause and derived actions, for every incident.
  • Operations report: monthly, with availability, incidents, utilization, recommendations.
  • SLA reports: availability and incidents, included in the Business package.
  • Dashboard: read access to live metrics, on request.
  • Failover test log: before go-live, then on an agreed schedule.
  • Documentation: architecture, access details and runbooks exist in writing and belong to you.

Capacity planning

Trend analysis and capacity planning with expansion recommendations

For capacity planning, our 24/7 monitoring tracks trends as well as faults, such as the utilization of compute, memory, storage and network. From these trends we derive expansion recommendations, which appear in the monthly operations report. That way you can plan and budget upgrades to your cluster in good time, based on measured data.

Trend detection

Proactive troubleshooting: we act on unusual trends as soon as they show up.

Backup monitoring

Backup monitoring: every backup job is checked

With backup monitoring, every backup job is checked. In regular restore tests we restore backups on a trial basis to confirm that they can actually be recovered. Backups run on infrastructure separate from the cluster, with defined retention periods; in the Business package they are stored, and restore-tested, at the second site NBG01.

A high availability cluster is not a backup.

Replication protects against hardware failure, but it also copies accidental deletions to every replica. Details: backup monitoring and restore testing.

Takeover

Handing over an existing cluster to managed operations

Already running a Proxmox cluster and want to hand off monitoring and on-call duty? After an audit of your environment, we connect the cluster to our monitoring and take over operations, updates and on-call duty. The step-by-step process is described under taking over existing Proxmox clusters.

FAQ

Frequently asked questions about operations and monitoring

Is my cluster monitored around the clock?

Yes. Our 24/7 monitoring covers all six layers of your cluster, from hardware to application. Alerts reach on-call duty through an escalation chain, at night, on weekends and on public holidays.

Which SLA do you offer?

One that we define together with you in the operations contract: availability target, response times by priority level, service hours, escalation paths and the recovery targets RTO and RPO. We do not promise blanket percentages without an architecture behind them.

How quickly do you respond to critical alerts?

Critical alerts reach on-call duty immediately and are handled right away, around the clock. We set specific response and recovery targets together with you in the operations contract.

What is in the monthly operations report?

Your cluster's availability, incidents and utilization, plus our recommendations, for example on adding resources. The Business package adds SLA reports covering availability and incidents.

Can I see the live metrics?

Yes. On request, we set up read access to a dashboard with live metrics, in addition to the monthly operations report.

Next step

Hand over operations and on-call duty

Tell us about your environment, your application and your availability target. We will show you what monitoring and escalation would look like for your cluster and what goes into the operations contract.

Hand over operations