As we enter a new year, it’s the perfect time to reflect on the data center management trends and challenges of 2020 and plan for what needs to be done in 2021 to improve uptime, increase efficiency, and boost productivity.
Start the year off right by creating a list of things you can accomplish in 2021 to optimize your data center sites. Then, see the results as the year goes by as you improve all facets of data center management from power monitoring to capacity planning to asset management.

Here are the top 10 New Year’s resolutions that should be on your list:

1. Find stranded power for all your racks. Data center managers often struggle with the complexity of accurately planning and managing power capacity. The traditional approach to power budgeting—derating the nameplate value to around 60% or 70%—is manual, estimated, largely inaccurate, and wastes money. Fortunately, there is a new way forward. You can now automate power capacity planning and increase rack power utilization by as much as 40% by leveraging a machine-learning algorithm that calculates power budget profiles for each of your device instances based upon how they are used in your environment. Find free power and confidently deploy it in your existing cabinets to defer building out unneeded capacity that can cost between $15,000 and $20,000 per cabinet.

2. Find and shut down your ghost servers. Did you know that an older server that is not ENERGY STAR compliant can consume 175 watts at idle, wasting space and money? These ghost servers need to be identified and consolidated to get the most out of your current power, space, and cooling capacity. If you don’t have the right tool, this may require a tedious manual process of looking up Excel files that store asset information, going to the data center to audit and verify the information, deploying workload tools that verify utilization rates, hunting down device owners, etc. Instead, leverage data center management software to easily run a ghost server report. Then, use that same software to manage the decommission process by knowing what assets and connections will be impacted, what additional capacity will be freed up, and automating the workflow process.

3. Clean up the cabling rat’s nest. Data center cable management is often considered a job reserved for network engineer, but the entire enterprise can be impacted by poor cabling practices that lead to unnecessarily difficult equipment installations, obstructed airflow that can overheat and damage equipment, and extended periods of troubleshooting and maintenance. Follow cable management best practices such as visually documenting your connections, precisely measuring cable lengths before installation, automatically validating connections, and tracking connectivity capacity with reports and dashboards.

4. Get a clear documentation of all your inventory. Modern data center environments are extremely complex with an overwhelming number of IT assets to monitor, and legacy management tools like Excel and Visio no longer get the job done. Make a resolution to upgrade your asset management capabilities by leveraging a Data Center CMDB to accurately track all IT equipment (servers, network, storage equipment) and supporting infrastructure assets (racks, rack PDUs, patch panels, structured cabling, patch cabling, UPS, busways, branch circuits). For complete asset management, be sure to track the detailed physical location of assets (e.g., site location, cabinet location, and exact U position), dimensional, weight, and physical data/power port information, physical relationships and connectivity, utilization and capacity of infrastructure resources (e.g., power, space, and cooling), and be able to remotely visualize your assets in rack elevation views with overlays of live power and temperature sensor readings. Get as granular as possible with your inventory documentation by tracking and monitoring spare parts such as hard drives, cards, memory modules, power supplies, patch cables, and any other component, even boxes of screws.

5. Get a clear documentation of all your power and/or network connections. Tracking your connections and port capacity has become critical for successful data center management. When you consider that a data center with 100 racks involves a complex system of over 70,000 port and cabling components, it becomes clear that you can’t afford to have poor documentation. Follow best practices and get a clear documentation of your connections to easily conduct impact analysis, find the ideal place to deploy new equipment, understand real-time available port and space capacity, document cabling installations, and know that your connections are compatible.

6. Implement grommets for bypass airflow. Using raised-floor grommets to eliminate bypass airflow improves cooling capacity and increases energy efficiency. If you are not already doing so, consider implementing grommets in your raised floor. The standard grommet is a rectangular shape cutout located 4 inches from the edge and centered on the panel. Placing grommets in the right place under the rack enables equipment to be easily deployed and upgraded without the need to replace panels.

7. Implement cold or hot aisle containment. Data center containment strategies offer many benefits including reduced energy consumption, increased cooling capacity, allowing for a stable supply temperature to IT equipment, enabling more power capacity available to IT equipment, increased uptime, and extending the lifecycle of IT assets. You can implement either hot aisle containment which encloses warm exhaust air from IT equipment and returns it back to cooling equipment or cold aisle containment which encloses the aisle where cold supply air is delivered, allowing the rest of the data center to become a hot-air return plenum. Both methods of containment offer unique advantages that you can benefit from.

8. Deploy environment sensors. Environmental monitoring is necessary to ensure the overall health of your data center sites, mitigate against threats of downtime, and improve efficiency. The most critical sensors to deploy in your data center are:

   1. Temperature. Monitoring for temperature helps you identify hot spots that can damage equipment and cause downtime or know if you are overcooling and overspending on energy costs. Follow ASHRAE guidelines to ensure proper sensor placement and that your equipment is in the recommended temperature ranges.

   2. Humidity. If your environment is too dry, static electricity can build up. Too humid, and equipment can corrode. In either direction, extreme humidity levels can damage equipment and as such you should be monitoring humidity per the ASHRAE guidelines.

   3. Airflow. Monitor airflow to help avoid hot spots and maintain a stable ambient temperature.

   4. Pressure. Monitor differences in air pressure to help identify air leaks that can cause inefficient cooling and hot spots.

   5. Vibration. Vibrations in the data center can potentially damage disk drives over time and cause downtime. As such, they should be monitored.

   6. Water. Early detection of water in the data center can give you enough time to prevent a potential disaster and downtime caused by air conditioning leakage, condensation, burst pipes, or local plumbing failures.

9. Implement power metering. Power meters provide critical data and insight into the utilization of your power distribution infrastructure to help ensure safe, efficient, and reliable operations. Data from power meters can enable you to increase uptime, improve capacity planning, manage existing capacity, report on Power Usage Effectiveness (PUE), reduce energy consumption and costs, and billback internal or external customers based on their consumption. Metered power infrastructure can include:

   1. Intelligent rack PDUs. There are several types of metering options with intelligent PDUs that distribute power within cabinets to IT equipment.

      1. Inlet metered. Metering at the PDU inlet helps determine power usage and available capacity at the rack.

      2. Outlet metered. Metering at the PDU outlet offers the same advantages of inlet metered PDUs, plus you can understand power consumption down to the device level.

      3. Circuit breaker metered. Know when a circuit breaker trips so you can reset it quickly.

      4. Outlet control. Remotely power on, power off, and power cycle individual outlets from any location.

   2. Bus drops and busway end feeds. Overhead power distribution systems can offer flexibility and ease of use. Depending on the model of your busway system, power meters may be at the bus drop above a rack or at the end feed of a row. Tap box meters measure the load at the outlet and end feed meters measure the load at the inlet of the end feed.

   3. Remote power panels (RPPs). RPPs distribute safe and reliable power from floor PDUs or other power sources directly to server cabinets.

   4. Floor PDUs. Floor PDUs are large, floor-mounted units that transform and distribute raw power feeds into lower capacity power feeds.

   5. Uninterruptible power supply (UPS). A UPS is a battery backup that ensures no power interruption occurs when there is a utility power failure.

   6. Building meters. Utility meters provide a measurement of the total facility power usage.

10. Deploy a modern DCIM tool. Remote data center management tools are no longer optional due to work-from-home employees and quarantine-related data center access restrictions. Data Center Infrastructure Management (DCIM) software is necessary in any data center management toolkit to bridge information across organizational silos and allow you to centrally manage all your resources and capacities in a single pane of glass to maintain uptime, improve efficiency of capacity utilization, and increase the productivity of people. Be sure to choose a modern, second-generation DCIM tool that offers:

    1. Zero-configuration analytics. Pre-built dashboards, reports, and interactive visual analytics that come out of the box without any tedious configuration effort.

    2. Data sharing and collaboration. Shared dashboards and team views that encourage information sharing and collaboration to break down organizational silos.

    3. Automation through integration. Out of the box ITSM connectors share data across disparate databases to save time and eliminate multiple manual data entry.

    4. Multi-vendor compatibility. Standards-based plug-ins allow you to manage all third-party equipment without being locked into specific vendors.

    5. Super-fast deployments. Easy deployment with little resources and effort that provides instant ROI.

    6. Scalability. Your solution should provide enterprise-class scalability that scales to handle millions of assets and sensors polling billions of data points daily without needing additional software licenses and server instances.

    7. Completeness of capabilities. Full-circle capabilities including asset, capacity, change, energy, power, environment, security, connectivity, visualization, and business intelligence and analytics.

Start with these resolutions to improve the management of your data center sites and by next year you will have more efficient capacity utilization, increased health of your environment and equipment, and boosted productivity of people.

Want to see for yourself how Sunbird’s second-generation DCIM software can help you implement these must-have tips and more? Take a free test drive today.