Cloud Services Proactive Performance Management

Service-centric, IaaS-based cloud services performance management across storage, compute, network and application domains.

Cloud Servcices Proactive Performance Management
Customer view for an IaaS offering showing the health of the the network connectivity, the virtual servers hosting the customer’s applications and their e2e latency.

The adoption of cloud computing is unprecedented due to its inherent flexibility and efficiency of resource utilization and dynamic responsiveness to changing demands on the underlying resources. Cloud services deliver simplicity for the end user but added complexity for the cloud provider that mu properly manage the availability and performance of these resources across storage, compute, network and application domains within a dynamic environment. So how does the cloud provider best manage their infrastructure-based cloud services (i.e. Infrastructure-as-a-Service)?

The key is to use a single solution that can not only manage the physical and virtual server systems but also the attached storage and local as well as wide area networking to build a complete picture of end-to-end performance and ensure the availability and reliability the cloud subscriber expects. In addition, linking the application to the network using flow-based data sources and deep packet inspection (DPI) technologies ensures that the cloud provider can understand what the infrastructure is doing; both the systems and the network, what the resource usage is and where its trending as well as identifying performance degradation anywhere with the multiple domains of IaaS so that operations staff and engineering can quickly rectify such degradation before impacting the end user experience and their SLA commitments.

Enhanced with a self-service dashboarding portal that allows a single-pane-of-glass view into the storage, compute, application and networking KPIs, the engineer is empowered to create their own view of the resources and services within their administrative domain, focus on the KPIs relevant to them, configure their own thresholds and define alerts for their mobile device.

Use Cases: 

See cloud service performance from virtual services to physical infrastructure to the end user

Visibility down to the end user is an important value-add, as no other service provider in the enterprise’s ecosystem will be able to monitor service levels to that degree of granularity.

Enable service visualization across multiple domains in the Cloud Provider

The service visualization capabilities that deliver this visibility can be used internally within the network operations center and engineering to help providers more effectively manage, plan and provision internal network resources and speed time to deploy new services.

Right-size internal compute, storage and network resources

Operationally, InfoVista allows the CSP’s internal engineering teams to use the same solution to right-size their physical infrastructure for capacity planning and provisioning purposes, allowing them to forecast their needs and minimize operational costs. This ability to right-size is a critical factor for the usage-based pricing models associated with cloud computing services.

Predict performance degradation of cloud services

Baselining capabilities plus forecasting that use historical data to detect significant change in service performance that may lead to SLA violations

Automate discovery physical and virtual compute and storage resources

Discover and monitor virtual entities such as VMware virtual machines, server hosts, data stores, clusters, UCS elements, SAN switches, storage arrays, controllers, disks, LUNs and disk volumes as well as data center network gear, such as load balancers, firewalls and high-end switches like Cisco’s Nexus, Juniper’s EX Series and Alactel-Lucent’s Omniswitch while managing the dynamic service-to-resource mapping.

Empower service operations’ engineers to rapidly detect and troubleshoot cloud service performance

With self-service analytics, historical reporting, user configurable thresholds and alerts and multiple views based on maps, lists and freehand drawings that provide the individual engineer complete control of their management domain

Deploy optional application performance management for added insight on cloud service performance

Deliver application dependency maps, application usage and performance visibility and triage performance degradation into application based, local data center network based or WAN based degradation.