MONITORING


FOCUS on POWER


Power made visible


Arange of smart hardware and software-basedpower usageproducts are starting to offer realways to‘map’ power in the data center


Power monitoring in any data center can be difficult. But what software and hardware solutions are coming onto the market to aid the data center operator as to the levels of power usage?


Purchasers of a modern data center power monitoring package should be able to find a set of common functions (see box, right). One data center supplier that clearly believes such software has a future is Emerson Network Power, which has combined its Avocent and Aperture purchases and systems into one data center infrastructure management solutions brand.


A core product here in power management terms is the DSView 3 PowerManagement suite. This offers the ability to monitor and measure IT energy consumption across individual PDUs, racks, rows of racks or entire data centers, by calculating and comparing cost across multiple levels (rack, row, data centers or companies).


Managers can set alerts to knowwhen the threshold is reached or exceeded and allow time tomake changes before there are actual problems, plus get that desirable“total reading”on power usage for the data center and set amaximumthreshold of power usage.


Also operating at multiple site scale is HP with its Data Center Smart Grid – a system that collects power and cooling information from across the data center, including both IT equipment, and power and cooling infrastructure. Data is collected by a multitude of sensors in real time and then aggregated to create its visual representation.


The idea is to communicate the same information to both the data center’s IT and facilities management for a more comprehensive approach to running the operation, leading to increased efficiency.


Other suppliers offering power management


software include nlyte, Servertech and Intel, whose 5500 series family of processors is being allied with a so-called dynamic Node Manager suite for better energy efficiency.


Wireless IP-based monitoring software Arch Rock’s Energy Optimizer uses wireless sensors to collect electrical, thermal, flowand air-pressure conditions in the data center, funnelled to an interface that compiles the data collected by the sensors and represents it visually based on the PUE metric and data center infrastructure efficiency (DciE) index.


Any number of sensors, each with a unique IP address, can be installed on power circuits, equipment racks, CRAC and CRAH units, on chillers and beneath raised floors. The variety of sensors includes outdoor devices that can help data center managers determine howmuch free cooling they can use in their particular location. The number of sensors depends on the level of granularity a customer desires in monitoring the facility, and comprise five types of sensor node, IP router and an interface application.


Not all power management software claims to offer such broad coverage. 1E, for instance, has an application designed to power down idle servers in the data center – the NightWatchman Server Edition, which identifies servers not doing usefulwork and puts them in the‘drowsy server’mode, running them at lowpower use without impacting their availability.


In theUS, Power Assure’s hostedMonitoring Software tracks and accumulates real-time data about power usage, temperature, humidity and application loads across the data center. Power usage and capacity ismeasured at the rack level and the cost of power is to be accurately assigned to each application, service or department, the developer says.


On the hardware side, supplier Raritan has a family of meters that monitor power use, temperature and humidity, its Dominion PX-3000 meters for use with standalone IT equipment, aswell as in-rack PDUs that do not have monitoring capabilities.


The meters feature built-in monitoring, security and networking capabilities, and


BOX 1: THINGS TO LOOK FOR IN AN EFFECTIVE POWER MANAGEMENT PRODUCT


• 24-hour CCTV monitoring


• Real-time measurement and modelling of server behaviour


• Customisable rack profile reporting features to show at rack level anomalous behaviour; for example, AC12 supply temperature, hot aisle average, temperature of cold aisle top, middle and bottom


• 2D thermal mapping functionality to identify hot and cold spots within the data center delivered by sensors placed at the top, middle and bottom of all relevant racks


• Overall data center thermal imaging


• 2D pressure mapping support to show where there may be too much or too little air pressure, which could cause air flow issues and/or hot and cold spots potentially impacting equipment


• Plenum differential pressure measurement (in inches or cms of water)


• Humidity below ASHRAE (www.ashrae.org) suggested ranges could indicate risk of static discharge, while if above, risk of condensation. Thus, there should be a way to capture relative humidity at the top of the cold aisle and air-handlers


• Individual branch power monitoring capability


• Individual circuit-level power monitoring charting kWh consumption


• PDU power monitoring


• Real-time power usage effectiveness (PUE) metrics


www.datacenterdynamics.com


13


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25