SolarWinds renamed its Hybrid Cloud Observability (HCO) platform to SolarWinds Observability Self-Hosted. The new name makes the deployment model clear: this edition stays on a platform you run yourself, on-premises or in a private cloud, and is distinct from the company's SaaS offering. The rename is more than marketing. It reflects one modern approach to observability while keeping the infrastructure architecture and the network, server and application monitoring that came before.
From monitoring to observability
Traditional monitoring tells you which system is down, using fixed checks such as up or down. That stops being enough in environments built from physical servers, virtualisation, containers and cloud services. Observability asks why it happened and where the root cause is, by analysing metrics, logs and traces together.
What it covers
The platform comes in Advanced and Essentials editions, and its capabilities fall into four areas.
- Network: traffic and flow analysis to see which users, applications and protocols use the most bandwidth. Dynamic path mapping shows the route a packet takes from source to destination, even across the internet and SaaS services such as Office 365 or Salesforce. Automated configuration backup, change monitoring and compliance checks are part of it.
- Infrastructure: monitoring of VMware vSphere, Microsoft Hyper-V and Nutanix AHV for capacity planning and for finding idle virtual machines, plus SAN and NAS monitoring (Dell EMC, HPE, NetApp) with forecasts of when space will run out.
- Applications and databases: monitoring of more than 1,200 applications and operating systems according to SolarWinds, including Windows and Linux distributions such as RHEL and Oracle Linux. Deep query and response-time analysis for SQL Server, Oracle, MySQL and PostgreSQL, and synthetic end-user monitoring to confirm that web portals really work.
- Hybrid and cloud: monitoring of IaaS, PaaS and SaaS services on AWS and Azure, and identification of over-provisioned or unused cloud resources to cut cost.
Analytics and AI
- Machine-learning anomaly detection: the system learns normal behaviour over time and flags unusual patterns even when no fixed threshold has been crossed.
- Less alert fatigue: correlated alerts are merged, so instead of hundreds of error messages the NOC or SOC gets one root cause.
- Drag-and-drop correlation: different metrics, such as server CPU, switch latency and database query speed, can be placed on one timeline to trace complex problems quickly.
Why I prefer a single platform
- One platform replaces separate tools for network, servers, Linux and cloud, which reduces licence and maintenance cost.
- Licensing is based on the number of nodes (devices) and not on ports or sensors, which is cheaper to scale in large environments.
- Unified data and end-to-end visibility help teams decide quickly whether a problem is in the network, the virtualisation layer or the application code, which lowers mean time to resolution.
- After the supply-chain incident that hit the company in 2020, SolarWinds rebuilt its software build process (its Next-Generation Build System) around zero-trust principles, with code produced in isolated, multi-stage environments.
What I learned
For organisations moving from traditional infrastructure toward hybrid and cloud environments, a single observability dashboard with AI-assisted correlation is a practical way to reduce human error and keep services stable, and it lets IT teams spend their time improving the infrastructure and not firefighting it.