Business owners seeking to protect their data with the strongest levels of security often choose to store their servers in data centers. After all, those facilities are georedundant, positioned in attractive geographical locations and secure.
Charged with the task of protecting the data of many businesses, data center managers should understand the importance of monitoring power at facilities in real time. After all, many businesses partner with data centers because they believe such facilities are the best option when it comes to safeguarding their data.
With this in mind, managers cannot afford to encounter any hiccups that can preclude their ability to deliver strong service to customers. And that’s precisely why it is imperative that managers do not overlook the fact that they need to actively monitor their data center power.
One easy way to do that is by investing in a reliable power distribution unit (PDU). Let’s take a look at three benefits such equipment affords:
• Diagnoses problems quickly. PDUs provide managers with pertinent data relating to irregular power consumption. Additionally, managers can set alarm thresholds that notify them when overloads are predicted to occur.
• Mitigates expensive repairs before they happen with an affordable solution. By proactively monitoring data center power, managers are able to predict areas where problems may occur and prevent them from happening before it is too late. Such behavior saves them from having to endure costly expenditures down the road to fix broken equipment.
• Establishes energy efficiency. PDUs enable data center managers to monitor energy usage in real time, identifying areas that are consuming more power and consolidating equipment in order to optimize the distribution data center power.
In order for a data center to be successful, it has to offer service that is second-to-none and have a long track record of reliability. With this in mind, data center managers cannot afford to leave any stone unturned and must employ modern technological solutions that ensure the continued deliverance of stellar service. Click here to learn more about how your data center power stands to benefit from PDUs.
InteropNet has been fully tested at the hot stage and is ready to roll for Las Vegas.
The 2014 InteropNet hot stage process has been wrapped up at the staging facility in Brisbane, CA. It’s amazing to see 20-plus cabinets go from being totally empty to fully utilized in such a short time.
It’s also just as amazing to know that now that it’s up and running, the InteropNet is going to be put into crates, shipped to the show site, reassembled and put into operation in time to support the show. This is a testament to the hard work and dedication of the volunteers and providers.
Creative minds from multiple vendors design and build InteropNet from scratch every year. Because the InteropNet is not staged in an actual data center, a number of challenges have to be overcome.
First is delivering enough power to the cabinets as 3-Phase power is not an effective option on a trade show floor. Second is closely monitoring the temperature, because where there is power there is heat. Hot spots are a challenge for anybody operating a high-density computing environment.
Server Technology joined the building process early to ensure that all the providers have both reliable and adequate amounts of power available for their devices. We work closely with the other infrastructure vendors such as Opengear and Cormant to ensure that our products are fully integrated to take advantage of requirements like power and environmental monitoring and remote rebooting of devices.
We monitor power and temperature usage and alarms very closely within this environment, right down to the outlet/device level. This scrutiny gives the NOC support staff and providers time to avoid problems before they occur and helps ensure overall uptime.
You can see the network in action with free, engineer-led tours of the InteropNet throughout the Expo. There are also behind-the-scenes views on Wednesday, April 2 and Thursday, April 3 at 11:00 am, 1:00 pm and 3:00 pm starting at the InteropNet Network Operations Center (NOC).
Explore troubleshooting techniques on a live production network in the workshop Network Troubleshooting Techniques Using the InteropNet. Top-rated Interop instructor Mike Pennacchi will walk through instrumentation, problem isolation and more. Register here.
Calvin Nicholson is Server Technology's Senior Director of Software and Firmware Development. Server Technology Inc. designs and manufactures intelligent power distribution products for enterprise datacenters. He was previously the Director of Product Marketing with ... View Full Bio
What if managing the datacenter were as simple as programming your DVR? Think about it – self balancing loads, automated response by the HVAC systems to temperature excursions, bandwidth on demand, uninterrupted computation despite a loss of power. Oh, that’s right. You already have that. Automation is great, isn’t it?
But what about when a hard drive fails? Or a tape drive gets jammed? A power supply goes on the fritz? Yes, the virtualization management systems can move the compute loads around, but eventually the hardware has to be repaired or replaced. Large scale facilities often use replaceable containers of compute infrastructure. Plug them in, let them run until they are unable to perform sufficiently, then recycle them. Other datacenters are not so fortunate. They have raised floor facilities that have racks deployed rather than containers. In some cases there are even hot or cold aisle containment systems that may limit access to the hardware.
Having a standardized approach to hardware design, a la “Open Compute” is a step towards being able to put robots to work in repairing the hardware, but it is my contention that a number of changes will be needed in both the hardware design and the design of the datacenter itself before we can begin turning robots loose to repair the infrastructure that is supporting our digital lifestyle. Meanwhile, there are a number of efforts already underway to work within the bounds of convention and still deliver an automated response to hardware issues. See the various links below to learn more.
Next month – where does rack level power fit into this?
1) Robotics Invade the Datacenter - http://www.informationweek.com/infrastructure/data-center/robotics-invade-the-datacenter/v/d-id/1112866 - Bill Kleyman video
2) Robots may run data centers, someday - By Joe McKendrick for Service Oriented | June 28, 2013 - http://www.zdnet.com/robots-may-run-data-centers-someday-7000017472/
“Kleyman does caution that "server hardware isn't quite ready to be handled by robotics," and the hardware has to be customized. Plus, many data centers still rely on cables, and it's hard to imagine a robot being able to thread and connect cables beneath raised floors or in ceilings. The rise of wireless connectivity could alleviate some of that.”
3) The Robot-Driven Data Center of Tomorrow - By: Bill Kleyman May 22nd, 2013
4) Data Center Robot in Action - http://www.datacenterknowledge.com/archives/2007/10/23/data-center-robot-in-action/ Tape library management with robots
5) Data Recovery Roundup: Robots in the Data Center, Backup-as-a-Service & Planning for Disaster - Megan Van Vlack on 17 Jan 2014 - http://subzeroblog.acronis.com/posts/data-recovery-roundup-robots-data-center-backup-service-planning-disaster
6) Roomba becomes data center robot - February 13, 2013 By Brian Benchoff -http://hackaday.com/2013/02/13/roomba-becomes-data-center-robot/
7) Toward Data center Self-Diagnosis Using a Mobile Robot – Lenchner, Mansley, et al http://www.research.ibm.com/people/l/lenchner/docs/icac2011_proofed.pdf
8) IBM Roomba-Based Robot Measures Data Centre Heat - June 3, 2013 by Peter Judge
“We currently have 11 production robots deployed in four continents, nine of which are at IBM data centres and two of which are used for customer engagements,” IBM researcher Jonathan Lenchner told TechWeekEurope. “We expect to have quite a few more robots deployed by the end of this year.”
The robots have recently been given the ability to scan RFID tags so they can add asset tracking to their role. Lenchner has been developing them for a couple of years, and has published papers on them (one of which is here).
9) Facebook Deploys Robots to Save Blu-ray From Extinction - By Cade Metz 02.06.14 http://www.wired.com/wiredenterprise/2014/02/facebook-robots/
One day, your Facebook photos will sit in the hands of robots. Behind the scenes at Mark Zuckerberg’s social networking giant, Facebook engineers have already built these robots, and one of them was on display last week in downtown San Jose, at a gathering of companies dedicated to exploring new technology inside the massive data centers that underpin the internet. This robot is little more than a mechanical arm — a device that moves up and down and side to side, grabbing things and carrying them from place to place — but it works entirely on its own, without human intervention, and it’s designed to grab things you’d never expect to find in the state-of-the-art computing centers operated by a company like Facebook.
Planning for growth in the data center starts with understanding where the capacity lies at the moment. For years, Server Technology's Sentry Power Manager (SPM) has provided the means for a detailed analysis of power usage in the data center by continual polling of in-rack power distribution units. Numerous features within the system allow for visualization, reporting, and trend analysis from which a data center manager can build an overall picture describing the current state of the data center. Furthermore, this can be extended into a plan for future growth including where to deploy new equipment and when to upgrade or extend the data center.
Over the next several months, my blog posts will focus on how some of our customers are using SPM today. We all know that operating a data center can feel like a full-time fire-drill. My hope is that by reading this blog, you will learn a few tips and tricks for making your analysis of the data center a more efficient process.
Some of the topics I plan to cover:
> How to identify the available places for new IT equipment.
> How to know when power availability will reach its limits.
> How to identify where efficiency can be improved.
> How to maintain redundancy within the data center.
Perhaps as you read, you will share even more of these tips in the comments.
“What I want to be when I grow up!” Career Week Presentation for UNR Engineering Students
I decided to jot down some thoughts for our blog while I was creating a career week presentation for a group of UNR Engineering students. In the mad rush to graduate and hopefully have a little fun while getting a degree most students fail to do any real career planning. After all they are going to school to become an Engineer so when they get out of school that’s what they will do. Right? I don’t have exact numbers but we all know that a large percentage of students in any discipline will work outside of the field that they received their degree in. For Engineers this often includes a number of different fields (see list below) but very often many of the skills that they learned in school are applied to these positions:
• Product Manager
• Sales Engineer
• Product Support
• Compliance / Compliance Engineer
• Manufacturing Engineer
• Test and Validation
• Patent Attorney
As part of their career planning there a number of things that they should do both while they are still in school and when they are out of school:
What to do -While still in school:
• #1 - Graduate
• School / Student Organizations
• Summer Jobs in your field
• Talk with Graduate Students
• Do your own career research
What to do- Once you are out of school:
• That’s when the real work starts
• Never Stop Learning
• More education/ MBA/ PE
• Toast Masters
• Invest in Yourself
• Know who your internal customers are!
• Plan your career track
Last but not least some skills / attributes that I think have value in the work place:
• Positive Attitude
• Willing to accept new challenges
• Good Communication Skills
• Ability to make complex problems/concepts simple
• Be a Thought Leader
• Even if you are the smartest person in the room don’t act like it!
• Do a lot less talking and a lot more listening!
• Don’t be afraid to make a decision
Today near the end of the “Great Recession” any advantage you have has a lot value! But as many of us know opportunity often sounds a lot like hard work.
See Ashton’s speech on You Tube for more details: http://www.youtube.com/watch?v=FNXwKGZHmDc
Smaller is Better.
How often do we hear that? Even when it’s a clear fact, it’s not a popular marketing phrase, but in the world of computing and electronics, it’s a hard truth.
That reality is what drove Server Technology to develop its own version of standard C13 and C19 outlets, which are 20% smaller than the off the shelf offerings. In addition to getting the ideal size of the outlet, we also designed in the other attributes especially suited for the data center environment, such as high retention, high temperature rating, specific shape and size the contact tab etc.
We’ve coined the term HDOT, High Density Outlet Technology for these outlets, and the products they’re incorporated in.
One Size Does Not Fit All.
In addition to packing more outlets in the same space, the other common request we hear from our customer is “Do you have that product, but with two more C19’s and two less C13’s ?” or “Can you move those C19’s to the other end?” To address the need for the exact right mix of outlets in the right place, we built our first line of HDOT PDU’s with configurable outlet modules. In this way the customer can specify from zero to fifteen C19’s, and twenty one to forty two C13’s, and have control over where those outlets are on the strip. This customizable approach allows for hundreds of combinations of outlets and their locations on the power strip. To easily specify the product, we created ‘Build Your Own PDU’ website, which walks you through a few simple steps to achieve the configuration.
Initial response from the HDOT product lines have been extremely encouraging, so expect to see many more HDOT products coming from Server Technology in 2014 and beyond.
If there is one word that can send shivers up the spine of an IT executive, that word is “downtime.” Aside from being a total nightmare to try and identify and fix, network downtime is also expensive. In fact, the average cost for a large enterprise that loses access to its data center is now climbing to about $212,200 per hour.
When assessing ways to maintain uptime in the data center, many companies will choose to move network operations to a geographically friendly location that is free of earthquakes, floods and tornadoes. Investing in network monitoring and fireproofing data centers also comes in handy to mitigate risk and ensure uptime. Consider this, though: about 40 percent of network outages can be attributed to unreliable power stations. A data center, in other words, can be located in the safest spot on Earth, but without a secure method of power distribution, it will still be prone to downtime.
If you want to maintain uptime in your data center, try looking into power distribution and monitoring technology. With such technology, it is possible to check up on input current monitoring and environmental monitoring such as the temperature and climate of your data center. You will gain the ability to see exactly where your power is being sent in your data center. Additionally, you can predict where future issues may arise on a PDU dashboard. This will send out alerts in real-time, and will diagnose problems as they occur so they can easily be fixed.
For more information on how you can gain access to power distribution monitoring in your data center, please click here.
CramIT! A Datacenter Blog
In today’s world of flexibility, you can have anything you want in your rack, so long as the form factor lets it fit inside the 19” wide space between the posts. Whether it’s switches, load balancers, blade enclosures or 1-u servers, hard drives or flash drives, they can be loaded in any order in the stack of your rack that you want. But when it comes to power, you pretty much have to wait for the PDU manufacturers to build the outlet combination you need to power your rack - until now.
With the advent of new modular Smart PDUs from Server Technology, you can pick and choose the outlet combinations you want to have in your power strip. Pick one of 14 different input cord types. Then choose top power input or bottom power input. Then choose from 6 different modular outlet groupings (12x C13, or mixed C19 and C13s, up to 5xC19 and 7xC13 per module), and mix-match three modules per strip, for up to 216 different combinations of power outlets. Then choose whether or not you want remote monitorability. Then pick your color (one of 5), and voila! You can have your PDU your way, just the way you like it.
1) Monitor Temperature and Humidity to ASHRAE TC 9.9 Standards- Slight increases in temperature can provide huge savings in cooling costs.
2) Employ a more efficient power strategy like 3-Phase 400/415 V power to the cabinet - Higher Power to the cabinet provides greater efficiency in your power distribution chain, servers run more efficient at higher voltages and this allows for greater densities and future growth
3) Employ blanking panels in the open spaces in the cabinet for better more efficient air flow
4) Implement a hot isle/ cold isle strategy and if you have already done this then implement some kind of hot isle containment
5) You cannot improve what you are not monitoring!
6) Monitor power at least at the cabinet level. This is critical for redundancy, ensuring uptime and for three phase load balancing at the RPP to ensure greater efficiency. This can also help you locate any stranded power capacity that you did not know you had.
7) Monitor power at the device level. This can help determine what devices are not doing any useful work and to also help determine where hot spots are located. Turning off “zombie” servers can provide a huge amount of power and cooling savings.
8) Monitor power usage of large storage devices like SAN’s or other large devices within your data center that consume a lot of power. Devices like these and other devices like large network switches often require specific power connections but only use a small portion of the connected power.
9) Shutting down unused devices- If you operate a lab environment of any environment not being utilized 24/7/365 then consider using a software tool that will allow you to schedule when devices can be turned off and back on again, as this provides a huge power savings
10) Consider other creative items to avoid problems like color coded cables for A and B in-feeds, a software tool that can check cabinet power redundancy or capacity planning via predictive trending, locking cables, remote power and environmental information access via Bluetooth to keep people out of the cabinets
Your data center is unlike any other on the market. You have
unique needs related to the space, size and budget of your operation. For this
reason, Data Center Managers cannot always assume a “one size fits all”
approach when it comes to purchasing data center and network infrastructure.
Conversely, optimizing for success in today’s data center requires custom solutions
capable of maximizing outlet density and space—and this all starts with the
right power distribution unit (PDU).
Up until recently, other brands of PDUs have proved
exceedingly challenging for IT managers looking to optimize space in the data
center. Due to the bulky PDUs and specifically the power outlets, which often
use excessive amounts of material, achieving maximum density has been a major
challenge up until this point.
Fortunately, Server Technology has liberated IT managers
from the confines of space-hogging PDUs through its latest release: the High Outlet
Density PDU and Build Your Own PDU configurator which grants IT managers the
ability to completely customize PDU design to your specifications. The process
starts with a 42 outlet Smart or Metered PDU frame and walks users along with a
step-by-step approach that covers the following options:
- Power inlet
- Connectivity options
- Outlet configuration
- Color selection
- Quantity (no minimum order)
With the ability to customize your PDU configuration, you
can maximize equipment density and also square footage without having to
sacrifice any data center power requirements. According to Server Technology
Director of Engineering Travis Irons, this is a brand new, never-before seen
initiative that will provide customers the benefit of a more efficient,
space-saving product. This all comes
without the added lead time and cost that other custom PDUs create.
“With six different outlet modules, you’ll be able to mix
and match where you want them in your PDU,” he said. “This will allow customers
to put the right type of outlet directly adjacent to the equipment.”
For more information about how you can implement a
customized PDU unit from Server Technology in your data center, please click