By Varma Kunaparaju,
For many organizations, the complexity of hybrid and dynamic environments can become overwhelming and hamper innovation. Adding to the challenges is the race to operationalize AI use cases while assuring governance, performance, and compliance. Finally, most organizations are concerned with the costs and productivity of their IT and AI investments and teams.
A new approach is needed to manage these large-scale, dynamic, and distributed environments—one that avoids human misses and helps predict and prevent issues. This approach must incorporate:
- Complete visibility across the hybrid and dynamic environment
- AI-powered analytics that constantly learn and relearn to keep pace with volume and speed of data
- Intelligent automation for immediate actions while keeping the human operators in control
This new approach forms the three tenets of OpsRamp’s vision for autonomous IT operations, delivering efficiency, agility, and resilience for hybrid cloud operations.
- Unified observability. OpsRamp delivers the ability to discover and monitor all the tools and technologies across a hybrid cloud environment in one place. OpsRamp supports more than 2,500 integrations.
- AI-powered analytics. OpsRamp applies analytics across all the telemetry data, makes sense of the alert floods, correlates upstream and downstream alerts to pinpoint the root cause of incidents, and understand trends and identify anomalies.
- Intelligent automation. OpsRamp uses policy-based automation to prioritize response and automate resolution of a number of routine tasks, such as configuration management, event and incident management, patching, and more. It automates escalation of the incident to the correct expert to resolve, including integration with ITSM and collaboration tools.
At this week’s HPE Discover 2024 in Las Vegas, we made four new product announcements to help further our long-term autonomous IT operations vision.
Operations copilot
OpsRamp’s new innovative operations copilot feature is a generative AI-based assistant that enables enterprises to detect, predict, and remediate problems more quickly by converting machine data into a human-actionable and human-friendly format. The operations assistant combines observability signal-specific AI models developed by OpsRamp with a GenAI conversational assistant to digest large datasets and provide insights in near-real time through intuitive and contextual dashboards generated on the fly.
Operations copilot is built on the foundation of unified telemetry and AI and ML-powered analytics from applications to infrastructure. OpsRamp uses its automation framework to respond to the recommendations suggested by the copilot, then takes action for orchestration and remediation.
Full-stack AI workload-to-infrastructure observability
We have extended OpsRamp’s full-stack observability to AI infrastructure and workloads, enabling IT teams to monitor their AI infrastructure in context of the AI workloads, correlated with the rest of their data center infrastructure. The integrations include NVIDIA GPUs and AI clusters, NVIDIA DGX Systems, NVIDIA Mellanox InfiniBand, and Spectrum ethernet switches. The metrics that are collected focus on availability, health, usage, performance, power consumption, and more. OpsRamp also provides integration with CrowdStrike for protection against misconfigurations and IAO detection. Customers can visualize the security posture of their AI infrastructure through the unified full-stack service map view in OpsRamp overlayed with security vulnerabilities. In addition, with HPE Ezmeral integration, devops, SREs, and operations teams can visualize AI workload-to-infrastructure performance, health, and utilization for a full-stack view of their AI deployments.
With this support for observability of NVIDIA infrastructure and AI workloads, OpsRamp can help organizations monitor and manage the performance of their distributed AI systems correlated with the rest of the hybrid cloud. This capability will be available as part of HPE’s Private Cloud for AI and with OpsRamp standalone SaaS service.
Application observability
OpsRamp has made significant investments in application workload observability, which is critical to understand how an application behaves in real time. The goal is to help IT and DevOps teams optimize application performance for both traditional and cloud-native apps. It helps answer many common questions like:
- Why is the application running slow?
- Why does the application have excessive HTTP errors?
- Why does the application have throughput issues?
It accelerates root cause analysis by incorporating logs, metrics, and traces to provide a deeper insight into the entire system along with an Apdex score. Our approach enables eBPF-based auto instrumentation of applications that generate telemetry signals (metrics, logs, traces) in OpenTelemetry (OTel) format. The collected data is exported to OpsRamp for deeper analytics, using the power of AI and ML to convert telemetry data into decisions.
Network observability
Autonomous IT operations requires 100% visibility of hybrid IT environments. With the introduction of full-stack network observability, OpsRamp has closed the visibility gap in network environments from the application to the edge. This includes support for observability of network flows, traces, logs, events, and metrics within a unified platform for application and infrastructure observability.
OpsRamp network observability supports full visibility of software-defined and virtual networks, storage area networks, and WiFi, LAN and WAN infrastructure. The new offering includes a suite of network performance management, network configuration management, and network topology management tools.
Final thoughts
In May 2024, HPE was recognized as a leader in the IDC MarketScape for Worldwide Multi-cloud Management with Automation 2024 Vendor Assessment. With these new announcements from OpsRamp, HPE continues to build on its market leadership and autonomous IT operations vision.
OpsRamp is available as a standalone SaaS service and as an integral part of HPE GreenLake Flex, Complete Care IT Operations, and managed services offerings.
HPE Experts
Hewlett Packard Enterprise
twitter.com/hpe
linkedin.com/company/hewlett-packard-enterprise
hpe.com
HPE_Experts
Our team of Hewlett Packard Enterprise experts helps you learn more about technology topics related to key industries and workloads.
