Platform for real-time storage performance analytics

Coraid® has announced its new platform for storage performance analytics – EtherCloud Insight™. EtherCloud Insight is the first enterprise storage performance platform that delivers visibility across the entire data path from the server, through multiple network paths, and into the distributed storage infrastructure. The platform enables data center operators to profile end-to-end performance for system design and to diagnose performance issues for problem resolution. By providing these capabilities in a single platform that can be accessed graphically or programmatically, EtherCloud Insight helps reduce wasteful storage over-provisioning, accelerate problem resolution and increase application uptime.

Lack of performance visibility in large shared-storage environments has become one of the top data center challenges, resulting in costly troubleshooting and downtime. Furthermore, for data centers supporting multitenant and diverse workloads, it has become nearly impossible to understand and engineer for the I/O needs of the many various applications and platforms. To meet storage requirements, enterprises and service providers have been forced to choose among costly options – over-provisioning every layer of the stack, creating multiple storage silos or installing separate legacy performance-analysis tools for each layer of the infrastructure.

Coraid’s new EtherCloud Insight redefines the model for storage performance analytics. Unlike legacy performance tools that require intrusive third-party instrumentation at multiple layers, EtherCloud Insight is seamlessly integrated into the storage infrastructure itself. EtherCloud Insight utilizes software sensors located in endpoints of the storage network, including Linux, VMware, Solaris, and Windows servers and Coraid’s storage appliances, to collect granular data on metrics including throughput, latency, IOPS, IO patterns, block size and queue depth. The platform collects data without an impact on the performance or reliability of the storage and the network. With these capabilities, the operator can rapidly analyze and troubleshoot performance issues before they result in downtime and ensure that the data centers’ resources are efficiently utilized.

“Data center operators have struggled to resolve storage performance issues in virtualized and cloud environments due to a lack of real-time, end-to-end visibility. EtherCloud Insight is the first platform that can deliver advanced performance analytics across the entire data path: server, network and storage layers,” said Coraid CEO Kevin Brown. “EtherCloud Insight delivers powerful, fine-grained analytics to reduce downtime, simplify performance troubleshooting and reduce over-provisioning in growing storage environments.”

“Coraid has won an impressive set of enterprise and cloud customers by combining enterprise storage features with an elastic scale-out architecture. EtherCloud Insight strengthens its differentiation in Ethernet storage by enabling detailed, real-time visibility into all of the key elements of storage performance,” said Ashish Nadkarni, research director, storage systems, at IDC. “Advancements of this type can help customers speed the transition from Fibre Channel storage networking to commodity Ethernet, thereby reducing costs and increasing flexibility.”

Integrated performance analytics are particularly vital in large-scale shared-storage environments such as cloud and enterprise data centers. EtherCloud Insight provides:
· End-to-End View: EtherCloud Insight is the first enterprise storage performance platform to deliver visibility across the entire data path from server to storage. This visibility, available without invasive third-party probes, enables operators to tune and troubleshoot multiple classes of storage from a single platform.
· Comprehensive View: A fine-grained, comprehensive view of storage workloads enables accurate analysis and troubleshooting. EtherCloud Insight allows administrators to monitor statistics at the granularity of a single drive, LUN or port. This data is available at the level of the operating system as well as at the network protocol layer.
· Explore-able View: Whether analyzing real-time or historical data, database views enable selection and isolation of system components to explore the data visually. The data aggregated from the sensor can be consumed through industry-standard tools such as Graphite or easily imported into third-party analytics platforms.