Data centers are rapidly adopting integrated infrastructure to run their businesses. However, the monitoring and management tools for these integrated components are not yet converged. Performance monitoring and capacity planning is done with multiple tools leading to sub-optimal resource management. Customers have to spend multiple days or weeks to install and configure typical agent-based solutions. Then they will have to train additional personnel to manage and maintain these multiple tools.
Cisco UCS Performance Manager provides visibility from a single console into Cisco Unified Computing System (UCS) components for performance monitoring and capacity planning. It provides data center administrators assurance for Cisco UCS and other integrated infrastructure implementations (e.g. FlexPod, Vblock, VSPEX) and ties application performance to physical and virtual infrastructure performance allowing IT staff to optimize resources and deliver better service levels to customers.
Cisco UCS Performance Manager was built in partnership with Zenoss. UCS Performance Manager is a virtual appliance designed for easy installation and configuration. UCS Performance Manager install and setup takes less than an hour, a departure from typical customer experience. Main customer requirements have been built into the product to provide key performance indicators out of the box. It provides a customizable dashboard for a quick view of the components of interest.
One of main constructs of UCS Performance Manager is user defined host groups. Customers can put related hosts in a group and UCS Performance Manager will automatically figure out the underlying infrastructure those hosts and applications are dependent on. This dynamic view provides the relationship and health of the hosts and related infrastructure for a quick glance of the status. Each of the devices and components can be clicked for more information about them.
From UCS perspective some of the frequent questions is how to find the bandwidth utilization of the server ports connected to servers, fiber channel uplinks going to SAN, Ethernet uplinks going to LAN and so on. UCS Performance Manager will provide this information in graphical and tabular form for easy consumption. Bandwidth utilization and health information overlay is applied on a UCS topology view for easy consumption.
By utilizing the views and trend graphs available customers can quickly identify any congestion points in the integrated infrastructure to prevent application performance degradation proactively. Administrators can provision additional resource to ease congestion or move load across the integrated infrastructure based on business needs.
With UCS Performance Manager in the data center customers will:
Have a better understanding of UCS integrated infrastructure at a component level
Maintain and provide high level of service with optimal resource allocation
Save time and resources with a single pane of glass for integrated infrastructure monitoring
We are presenting an overview of UCS Performance Manager at the Cisco booth at VMworld next week. If you are attending the event stop by for a demo and live Q&A. Get a first look on the show floor at the Cisco booth #1217.
While at VMworld attend the theater presentation and break out sessions for more information.
1. “Performance and Capacity Management for Cisco Converged Infrastructures”
Tuesday August 26, 2:00pm, Cisco booth #1217
2. “Management and Automation for UCS Integrated Infrastructure”
Tuesday August 26, 3:30pm, Room 3022 Moscone West
The Cisco UCS® C240 M3 Rack Server captured the number-one spot for overall price/performance on the TPC-H benchmark at the 1000GB scale factor with a price/performance ratio of $0.73 USD per QphH@1000 GB and demonstrated 304,362 queries per hour (QphH@1000GB), making it the fastest two-socket server running Microsoft SQL Server 2014.
The TPC-H benchmark evaluates a composite performance metric (QphH@size) and a price/performance metric ($/QphH@size) that measure the performance of various decision support systems by running sets of queries against a standard database under controlled conditions. As tested, the benchmark configuration consisted of a Cisco UCS® C240 M3 Rack Server equipped with 768 GB of memory and two 2 Intel Xeon E5-2690 v2 processors. The system ran Microsoft SQL Server 2014 Enterprise Edition and Windows Server 2012 R2 Standard Edition. Check out the Performance Brief for additional information on the benchmark configuration. The detailed official benchmark disclosure report is available at the TPC Results Highlights Website.
Some of the key highlights of Cisco’s TPC-H Benchmark results are:
Best Price/Performance: The Cisco UCS® C240 M3 Rack Server captured the number-one price/performance spot on the TPC-H benchmark at the 1000-B scale factor with a price/performance ratio of $0.73 USD per QphH@1000GB. This result beats 8-socket HP ProLiant DL980 G7 runningMicrosoft SQL Server at 219,887 QphH and $ 1.86 USD/QphH@1000GB by 60 percent.
Best Two-Socket Server Performance for Microsoft SQL Server 2014 Cisco demonstrates 304,362 queries per hour (QphH@1000 GB), making it the fastest two-socket server running Microsoft SQL Server.
Steve Jobs is arguably the most amazing innovator of our times. I recently read some of his thoughts on innovation. His statement “Innovation distinguishes between a leader and a follower,” caused me to reflect upon my eight-year association with data virtualization, and consider who in the IT analyst community have been the innovative leaders.
The role call of top analysts doing innovative work continues with Noel Yuhanna of Forrester who wrote the analyst community’s first research paper on data virtualization in January 2006, Information Fabric: Enterprise Data Virtualization.
Gartner’s Ted Friedman and Mark A. Beyer, and more recently Merv Adrian, Roxane Edjlali, Mei Selvage, Svetlana Sicular and Eric Thoo, have been both descriptive and proscriptive about the use of data virtualization as a data integration delivery method, a data service enabler and a key component in what Gartner calls the Logical Data Warehouse.
Further, there have been myriad analysts who have amazing contributions.
The learned trio of Dr. Barry Devlin, Dr. Robin Bloor, and Dr. Richard Hackathorn have pushed the art of the possible.
While analyst / practitioners such as Jill Dyche, Mike Ferguson, Rick Sherman, Steve Dine, Evan Levy, David Loshin and William McKnight, via their hands-on client work, have “kept data virtualization grounded on reality street,” to quote Mike Ferguson.
And let’s not forget the Massachusetts’ Waynes — Wayne Eckerson formerly of TDWI and Wayne Kernochan, author of the eponymous Thoughts From a Software IT Analyst blog. Their voices and insights have proven invaluable.
To quote Gene Rodenberry, “It isn’t all over; everything has not been invented; the human adventure is just beginning.” The same is true for data virtualization. So I look forward to more great insights from these innovators, as well as a new generation led by Puni Rajah of Canalys and Vernon Turner of IDC.
To see Rick van der Lans and Barry Devlin on stage and gain even more insights from the 2014 Data Virtualization Leadership Award winners, join us at Data Virtualization Day 2014 on October 1 in New York City.
Watch for a sneak peek of Data Virtualization Day 2014.
To learn more about Cisco Data Virtualization, check out our page.
More than 5 years ago, Cisco introduced its first Nexus 2000 Fabric Extender (FEX) into the market. This broad portfolio has enabled over 15,000 customers to seamlessly add network access port density across server racks without adding the inherent management complexity that comes with adding more boxes into your architecture to meet scaling requirements which ultimately translates into lower capex and opex for the business.
Today marks the next evolution for this portfolio as we introduce the Nexus 2300 platform -- the 3rd generation Fabric Extender family. Based on the extensive innovations you have come to know from the Nexus 2100 and 2200 platforms, Nexus 2300 Fabric Extenders expand on these capabilities with:
Larger buffers to absorb bursts of traffic for a wide variety of workloads such as multicast feeds, voice traffic, video traffic, and healthcare applications
Unified Ports support enabling a flexible LAN and SAN deployment through support for Ethernet, Fiber Channel and Fiber Channel over Ethernet connectivity
Support for Cisco’s 40G BiDi optics simplifying migration 10 to 40 Gigabit Ethernet speeds while reusing existing 10G cabling
Additional versatile TCAM which can be used for:
Advanced features such as ACL classifications and QOS
Hardware-capable local flow redirect for architectures that require intra-rack traffic to reduce bandwidth
Put all of these together with the single point of management, policy enforcement, zero-touch provisioning installation and automatic configuration that is available across all Nexus 2000 Fabric Extenders and you further benefit from a more simplified and flexible network design that helps you commission and decommission server racks faster, simplify operations, and support varying workload requirements.
Now let’s look at the first member of the Nexus 2300 platform, the Nexus 2348UPQ. Priced at $9500 (US List), almost the same price as current 1Gbps FEXs, the Nexus 2348UPQ supports 48 1/10 Gigabit Ethernet ports. You can add further 10 G ports when you split the 40G ports for up to 64 10G ports – perfect for data centers that are migrating their servers from 1 to 10G network connectivity. The 6 on-board 40Gbps ports support Cisco’s BiDi optics so that as you migrate from 10G to 40G as your uplink speed, you can reuse your existing 10G cabling – helping you save on the costs of re-cabling your network and get it upgraded faster! This new fabric extender can be deployed in conjunction with Nexus 5500, 5600, and 6000 parent switches and with the Nexus 7000 and 9000 Series in the future.
In summary, Cisco’s Fabric Extender portfolio has seen tremendous traction in the market as the many benefits of this architecture ultimately help create a cost-effective, flexible and simplified approach to building a data center network. The Nexus 2300 Series Fabric Extender, with the Nexus 2348UPQ, continues to deliver on these same principles and further expands on the promise of helping simplify network deployment and operations while ensuring the data center network is ready to support varying application needs. I invite you to learn more about the Nexus 2348UPQ and other Nexus 2000 Series Fabric Extenders at www.cisco.com/go/nexus2000.
In this week’s episode of Engineers Unplugged, David Hartman and Tim Cerling discuss Fast Track 4.0 Solutions, which promote fast and efficient deployment of private clouds with Cisco, EMC, and Microsoft solutions.
How many engineers does it take to straighten a whiteboard? (Answer: 5) Behind the scenes on #EngineersUnplugged
If you would like to become Internet Famous, and strut your unicorn talents, join us for our next filming session at VMworld 2014. Tweet me for details!
This is Engineers Unplugged, where technologists talk to each other the way they know best, with a whiteboard. The rules are simple:
Episodes will publish weekly (or as close to it as we can manage)