Automated infrastructure management, DCIM and more advanced capabilities aim to eliminate the blind spot in full-network management.
by Patrick McLaughlin
The management and administration of a network's physical layer--primarily the structured cabling system--has advanced technologically over several years to the point at which latest-available technologies stake a claim to erasing the physical layer's status as a network-management blind spot. As we have pointed out previously (see "Expanding the realm of cabling-system administration," May 2013), managing a structured cabling system may begin with a standards-based labeling scheme, but it does not end there. Today's administration capabilities empower cabling and network managers to have a comprehensive view of the physical layer, and also to view that layer together with (not separate from) the network's higher layers.
During a web seminar hosted by Cabling Installation & Maintenance in July, speakers addressed this concept while simultaneously discussing some of the specific technologies available today to achieve higher-level cabling-system administration. The seminar was titled "Cabling System Administration and Management" and included presentations delivered by James Cerwinski, director of software for Raritan (www.raritandcim.com) and by Blake Van Scoy, product specialist with CommScope (www.commscope.com).
The current reality
Cerwinski addressed the topic of data center infrastructure management (DCIM), providing a high-level overview of such a system's capabilities. Focusing primarily on its use in data center environments, Cerwinski effectively made the case for DCIM's use by describing, in detail that is all-too-familiar for many data center managers, typical scenarios in these facilities. Addressing these individuals directly, he stated, "You as data center professionals are often acting in hero mode. You're being asked to do more with less. You're asked to improve uptime and SLAs [service level agreements], to optimize with existing resources. You're being asked to do this with fewer people, with less power, and with less equipment. This was a difficult task when data centers were simple and not so dense. As years have gone on, with data center consolidation, the boom of mobile applications and the need for hosted computing applications, data centers have become much denser and more complicated."
He then presented the scenario that clinches the DCIM value proposition: "You as data center professionals are being asked to do this with the same spreadsheets and drawings you've always used to manage your data center. A typical data center operation I review still manages enterprise-class data centers with a collection of spreadsheets that contains data that is impossible to correlate. All that data cannot possibly be correlated together to give that data center professional the information they need at their fingertips when they need it.
"This lack of an integrated system to provide that information, at their fingertips, with visualization, causes data center professionals to do what they have always been doing--to manually walk the data center. If they need to add a piece of equipment, they look at their spreadsheet to see where they might have available space. They look at a different spreadsheet to see where they might have available network connections and power connections. And they'll visually walk the data center to confirm the information in the spreadhseets--because that information in the spreadsheets can be unreliable. After walking the data center in hopes of seeing space where they can place a piece of equipment or make connections, they go back to implement the change before someone else reserves and takes that space."
What DCIM does
While stating his time-limited presentation could not go into great detail on DCIM, Cerwinski explained that such a software package "is a better way forward," because it enables a data center manager to "bring all the data in your disparate Excel spreadsheets into one relationship database, where that data can be automatically correlated, and validated information presented to every user, when needed, with the click of a mouse. DCIM ultimately is trying to achieve the objective laid out--increase uptime while being as efficient as possible with your capacity utilization."
A full DCIM system provides the user with management capabilities for every significant aspect of data center administration--capacity management, change management, power management, environment management, energy management and asset management. Its benefits are many, Cerwinski explained during his presentation, and paid specific attention to the ability to "reduce operational expenditures by automating manual and time-consuming processes," and the ability to "reduce and defer capital expenditures by allowing you to maximize utilization of your resources including power and connection."
CommScope's Van Scoy echoed several of those sentiments during his presentation, which focused on automated management of cabling systems and included mention of the company's imVision Infrastructure Management system. Van Scoy addressed both the data center facility and the enterprise environment in his presentation, explaining how automated systems can be beneficial to each.
Van Scoy described the enterprise network transformation that has given rise to the need for automated infrastructure management, and also has paralleled the development and capabilities of such management systems. He explained that historically, "Most networks were single-purpose--voice, data, and video and the three do not cross." Acknowledging that is still the case for some networks, he said in such environments, "There are not many devices, and they tend to be static once placed." These types of networks also are characterized by segregated operations, and those groups within the user organization generally do not intermingle. Importantly, in these environments, changes are manual processes.
Today, on the other hand, "Many organizations have multipurpose, dynamic networks," Van Scoy said. "Individuals come in connecting to WiFi, hotspots, or use hot-desking," in which a single desktop or user station is open for use by any authorized network user. In these environments, "everything is integrated," he added. "Take for example a hospital in which the pharmaceutical refrigerators are on the network." In such an environment, he said, "Groups that had not worked together previously now are working together. However, there are not more people than there had been before … So we have to do more things with fewer people," he said, akin got Cerwinski's description of the modern data center environment. "How do we automate that?"
As these networks have evolved from being single-use to multi-use, the management of network infrastructure has followed a similar path. Van Scoy explained, "Network management systems have come online. Asset management has become very important, especially since Sarbanes Oxley's introduction in the U.S. and other laws in other countries about controlling assets."
Capacity planning is also more imperative than before, as he illustrated: "Why buy a switch today that you won't need for six months? Or, what if your capital process takes six months? You need to plan today for what you will need later." He emphasized the growing importance of energy management to any facility--data center or enterprise--over the past few years.
A parallel exists between building systems and network systems, Van Scoy explained. Just as voice, data and video networks previously operated independently but now have converged, building systems like energy, power, communications networks and more are increasingly interactive with each other. Like Cerwinski, he pointed to DCIM as a means for taking a wholesale view at a facility's systems. Van Scoy specifically addressed how and where a structured cabling system fits in when he said, "Without knowledge of the physical layer, it is very difficult to have a good DCIM system."
Visibility into a network's physical layer, such as that provided by an automated infrastructure management system like CommScope's imVision, eliminates the blind spot that Layer 1 traditionally has been in network management. Van Scoy laid out the example of keeping a network operational at "five 9s"--99.999 percent of the time. "If a business support team spends 20 minutes trying to understand where the physical layer is, you're never going to be able to meet SLAs. Understanding the physical layer and being able to rule it out as your first troubleshooting step will quickly help you get onto the next network layer."
Additionally, typical network-management systems can see only so far down the stack. Specifically, as Van Scoy explained, "Many Layer 3 devices can tell you what switch port a device is connected to. It is great to know where the switch port is, but you need to know where that switch port connects out into the environment."
Until recent technological developments came to fruition, connection visibility ended at the switch port. Today solutions like imVision enable such visibility into the physical layer. The system's software package along with iPatch hardware--intelligent panels and shells, as well as a controller--enable a network manager to see a connection point all the way to the user cord.
Curing the network blind spot that has been the physical layer was exactly the intention when TE Connectivity's (www.te.com) Quareo system was developed. Gene Malone, product manager for TE's managed connectivity systems, explained in an interview with us that Quareo is "a system within the umbrella of managed connectivity" that the company offers.
Malone expanded on the troubleshooting scenario: "Sixty-five to seventy percent of network outages have to do with something on the physical layer. A lot of time is spent figuring out where the problem is." It is estimated that about 90 percent of troubleshooting time is spent diagnosing, identifying and/or locating the problem, with 10 percent spent actually resolving it. The Quareo solution, he explained, drives the diagnose/identify/locate time down significantly.
TE Connectivity describes Quareo as a Connection Point Identification (CPID) system. Each piece of connectivity in the system has a built-in microchip that carries information about the connection. Malone pointed out that each chip has approximately 20 data points, including the media type (copper or fiber), media construction (e.g. shielded, unshielded, multimode), the circuit's length and, among several others, the number of insertions the connection has experienced since being turned up.
Quareo integrates with TE Connectivity's Infrastructure Configuration Manager (ICM) software. As the company explains, "ICM is what enables the automated, accurate, real-time physical layer management of TE's managed connectivity." The software interfaces with Quareo, as well as with the company's AMPtrac automated infrastructure management system "to provide a live graphical view of the physical layer topology, including Layer 2 and Layer 3 devices. ICM explores, discovers and maps all the connections of the network, identifies components by name, and automatically records and provides notification of any changes to the physical layer cabling connections--all in real time for immediate response, increased network performance, and reduced downtime."
The complete visibility of a network's Layer 1--the physical layer--is Quareo's piece de resistance. Malone explained that this capability "allows a level of reporting not available using logical tools." One way of viewing the system is that it bridges Layer 1 with all other network layers in terms of their ability to be managed. A Quareo-managed physical layer coexists with whatever network-management system a user already has in place.
The system has benefits for data centers as well as office-building-type networks, and TE Connectivity promotes it to both network types. In data centers, Quareo is seeing increasing uptake from colocation facilities. Its ability to provide extensive detail on port connectivity enables these facilities to report on when circuits were connected and disconnected, down to the minute, for billing purposes. Additionally, TE Connectivity reports, service providers are using Quareo not only to live up to SLAs, but also to analyze the efficiencies of their networks and their personnel. The system allows these providers to examine which buildings, rows, or racks are requiring the most time and attention, for example.
Earlier this year the company released an ICM mobile app. In conjunction with an active ICM system, including the Quareo system, the app provides "unprecedented management, details and control of the information transport system," the company explained when the app was introduced. It features several work-order functions including work-order selection listing, work-order synchronization, work-order/LED activation controls, and work-order step selector/manager. Additionally, the ICM mobile app includes auto-detect and manual completion indicators, response-attachment handling and incident-response logging.
Networks of all types have become increasingly complex and the tools to manage them more sophisticated. Current technologies allow integration of physical-layer management with other network layers that was not practical a few years ago. ::
Patrick McLaughlin is our chief editor.
View Archived CIM Issues