Cisco UCS Central Software Plus Splunk Enterprise: End-to-End Operation Visibility for Multiple Cisco UCS Domains with Machine-Generated Data
PDF(168.0 KB) View with Adobe Reader on a variety of devices
Updated:November 21, 2012
Gain real-time operation visibility into multiple Cisco Unified Computing System™ (Cisco UCS®) domains with Splunk Enterprise integrated with Cisco UCS Central Software, which together can identify and resolve problems faster, proactively monitor systems and infrastructure, track key performance indicators, and understand trends and patterns of activity and behavior.
Cisco Unified Computing System and Cisco UCS Manager
Cisco UCS is the first data center platform that integrates industry-standard x86-architecture Cisco
® servers with networking and storage access into a single converged, computing, networking, and storage system. Cisco UCS Manager provides a single point of management for each Cisco UCS domain of up to 160 servers and associated infrastructure. Using a policy-based approach to server provisioning based on service profile templates, Cisco UCS Manager is used by administrators to quickly reproduce existing physical configurations, including I/O, firmware, and settings. A role-based access control (RBAC) model helps ensure the security of system configurations.
Cisco UCS Central Software
Cisco UCS Central Software extends Cisco UCS Manager features across multiple Cisco UCS domains. It allows companies to manage unified computing environments on a global scale, putting computing capacity close to users while managing infrastructure with policies defined centrally. With Cisco UCS Central Software, helping ensure global policy compliance is much easier, with subject-matter experts able to choose the resource pools and policies that must be enforced globally or locally. Cisco UCS service profiles can be moved between geographical locations with drag-and-drop simplicity to enable fast deployment of infrastructure as needed to support business workloads.
Splunk Enterprise for Cisco UCS
Cisco UCS environments are dense computing environments that often run hundreds of applications, different operating systems, and hypervisors. Data such as faults, events, power consumption, temperature, and performance can be analyzed to understand system behavior, user experience, application response times, and much more. The collection and cataloging of this massive amount of data together with data from applications, operating systems, hypervisors, and other devices in the environment is a Big Data challenge.
Splunk Enterprise is a scalable and highly versatile engine for indexing and harnessing the intelligence of machine-generated Big Data across a virtualized infrastructure. Splunk lets you:
• Centrally monitor, implement alerts for, create reports for, and analyze metrics, logs, and events in real time across the entire physical and virtual stack
• Correlate and connect events across every level and technology with Splunk's powerful search language
• Proactively detect performance problems and prevent them from affecting end users
• Determine root causes of outages and performance problems up to 70 percent faster
• Retain transient data from every element for trending, historical analysis, security, and compliance
• Flexibly address reporting and operations analytic requirements such as capacity planning, use analyses, and asset reporting in the continuously changing virtual environment
• Scale to handle Big Data problems faced by the largest data centers with Splunk's unique MapReduce-based, schema-less technology
The Splunk App for Cisco UCS works with Cisco UCS Manager and Cisco UCS Central Software to provide the power and flexibility of Splunk in an experience tailored for Cisco UCS technologies (Figure 1). Splunk for Cisco UCS:
• Provides real-time and historical visibility centrally across an entire Cisco UCS deployment
• Helps correlate Cisco UCS performance, fault, and events data with user, application, and hypervisor data to analyze, prevent, and fix problems
• Helps proactively monitor every Cisco UCS environment by providing analytics such as available capacity, trending of faults over time, and tracking of power and cooling costs
Figure 1. Displaying Server State, Availability, Model, and Available Capacity Across Cisco UCS Domains
The Splunk App for Cisco UCS uses the XML API to gather inventory, performance, and fault information. It also uses information sent over syslog, including configuration and state changes and authentication, authorization, and accounting (AAA) information. The Splunk App for Cisco UCS includes views such as:
• Central operations view across multiple Cisco UCS domains
• Trending of faults over time, by cause and by affected assets
• Trending of power consumption, cooling requirements, and networking throughput over time
• Authentication tracking
Correlation of Cisco UCS data with data from other technology layers yields critical insights that can help resolve user- or application-level problems and generate operations analytics such as information tracking the impact of changes across technology tiers. When problems arise at the user or application layer, the capability to correlate events from the underlying hardware through the virtualization stack all the way to the individual application or session can dramatically reduce problem resolution times. With the Splunk App for Cisco UCS, you can now visualize Cisco UCS data in Splunk alongside other operational data such as hypervisor, OS, and application data and use the Big Data analytics capability provided by Splunk to sift through massive volumes of data for faster troubleshooting, root-cause analysis, proactive monitoring, and operations intelligence.
Download Splunk for free. You get a Splunk Enterprise license for 60 days, and you can index up to 500 MB of data per day. After 60 days, or any time before then, you can convert to a perpetual Free license or purchase an Enterprise license by contacting firstname.lastname@example.org.