Thank you for contacting us. We will get back to you shortly.
June 27, 2011 - by badnima
Introspection is defined as the “contemplation of one’s own thoughts, impressions and feelings”, but if you’re an application running in the cloud, introspection reflects “why” as much as “how”.
Moving from managed infrastructure to cloud-based solution often involves a big tradeoff in transparency - transparency in the control layers, visibility into how systems are setup, how well servers are running, and where performance problems are occurring. This has long been accepted as the price we pay to utilize automation and abstractions that remove complexity from the cloud.
But it doesn’t have to be that way.
Cloud Analytics are a standard feature in SmartDataCenter 6. At the core, cloud analytics provide scalable, unified instrumentation and post-processing for deep observability within systems and applications in the cloud.
Built using DTrace hooks throughout the stack, SmartDataCenter collects and presents real-time and historical data to perform analyses across individual cloud nodes as well as applications distributed across the cloud. The user experience is very interactive. While looking at some data, you may instruct the system, through the Cloud Operator Portal or Public API, to collect additional data from cloud nodes for immediate display and analysis.
The primary metrics are direct and linear measurements of performance, such as workload analyses and resource monitoring. Workload analyses are a unique set of latency-based analytics that identify, locate, and quantify top-to-bottom performance issues. The secondary metrics explain the root cause of an issue through decomposition (a.k.a., break-down) and predicate (a.k.a., drill-down) of the primary metrics for each process, node, request, or system component of your choosing.
To learn more about Cloud Analytics, take a moment to watch the following videos.