Green Notes

It seems like the most simple logical solution to energy conservation-powering down servers not in use. After all not all servers are used all the time. For example servers that perform only a single predictable function such as for calculating the employee pay are used only on certain days in a month. Or those used for special testing or running backups and so on. It makes sense that they power themselves down when not in use just like in commonplace electronic gadgets.

The Alliance to Save Energy, Kelton Research and 1E, a software IT company, found through surveys of server managers and analysis of industry data that out of the world’s 44 million servers, roughly 4.7 million of them are idle and not doing necessary work.

“Contrary to popular belief, one of the largest causes of energy and IT operational waste in data centers are servers that are simply not being used. The savings from decommissioning non-productive servers cannot be ignored. Organisations need better information on server efficiency and more effective ongoing server energy management,” comments Sumir Karayi, CEO, 1E.

The key problem highlighted in the study is that over eight in ten (83%) of IT managers admit that they do not have an adequate grasp of server utilization . That is they have no idea which of the servers are actually in use and which are not at a given point of time. So to be on the safe side and avoid any unwanted interruptions while running applications, they supply power to all servers uniformly even those that are not working.

But not all IT managers approve of powering down servers. The old school of thought seems to be that servers should run 24/7 for fear of ruining the hardware and also of getting a bad rap sheet from clients. Others dread the difficulty of getting a server back to an operational state once it’s turned back on after powering down.

Ted Samson in infoworld interviewed several experts on this issue:

Infrastructure technologist at HP, said powering down servers is completely safe. “It’s not at all bad for the server. It’s something we do to electronic devices all the time. It can handle it from a hardware perspective,” said Baker.

But Brad McCredie, an IBM fellow for the Systems and Technology Group, wasn’t quite so sanguine. He explained that, technically speaking, powering off and on any kind of computer can have a detrimental effect over time.

“[Temperature cycling] is a well-established failure mechanism and a stress on components,” McCredie pointed out. “What it really comes down to is all these things — chips soldered on modules, soldered on boards and connectors — that expand and contract when they heat and cool…. When they all contract and expand at different rates, they can fail. That’s ultimately the bad thing with power cycling,” he said.

Mark Monroe, director of sustainable computing at Sun, suggested that machines can handle being shut down a finite number of times. Arguably, the number is large enough for regular power cycling over an extended period of time. “Most server vendors today say they’ll support a certain number of cycles of powering things on and off,” Monroe said. “I believe most of the server vendors would say [the number] is in the hundreds as opposed to the thousands.”

From the above survey it is safe to assume that for the most part , datacenters armed with the burden of reducing power costs are willing to at least take a look at this controversial method in moderation.

There have been a host of products in the market recently to tackle efficient power downs without mess-ups. IE launched its power management software for servers that use smart algorithms to figure out which servers are doing work and the energy they are using. According to IE its “NightWatchman Server Edition” can power down servers into a drowsy state, cutting their energy use by 12 percent. IBM has a product called Active Energy Manager that features advanced energy control options designed to boost performance per watt by slowing processor clock speed or even putting processors in “nap” mode when not in use. There are numerous others following the same trend including VMware that offers a functionality in VMware Infrastructure 3.5 called Distributed Power Management that powers servers on and off according to demand level.

Not all servers are automatic candidates for shutdowns however. Servers dedicated to specific scheduled computing that occurs only at certain times of the month or wanes at certain pinpointed times during the day are ideal targets for powering down. For more expert consultation on powering down processes and other methods of energy saves for your data center, contact us at GreenRack Sysems/408-2138144.

While the high tech world buzzes with non stop talk about virtualization and other power monitoring technologies to increase data center efficiency, improvement in basic storage can go a long way to green up your facility. On an average 10 to 40 percent of energy cost in data centers is used to run storage. Yes, managing storage can be expensive especially in a situation like data centers which struggles to cope with excess of data resulting from global initiatives, web commerce, rapidly enhanced computing expectations and security pressures. If businesses do not address their storage budget issues and reduce their data footprint, the problem is likely to blow out of proportion in this era of faster, more ever increasing computing. To keep up with this massive growth, organizations must look at smart storage capacity solutions.

SNIA, Storage Networking Industry Association is a non profit association made up of about 400 companies dedicated to knowledge exchange of end-to-end storage solutions and development of standards in information management products. Some of the energy efficient storage solutions that it recommends are thin provisioning, multi-use backup, virtual clones, de-duplication, compression, using RAID 5 or RAID 6 disk arrays instead of the higher-level standards and smarter coupling of technology and the actual needs.

Most industry experts recommend against indulging in highest grade equipment and to stick to what is actually required instead. Multi purpose, elaborate storage solutions not only cost more but most of the times you may not even be in need of them. So you might be powering them for doing things that you may not even need. So stick with RAID 5 if that’s all you will be using. While Solid State Disk translates to an enhanced performance level, it also costs an arm and a leg. So data that is not critical can be allocated to lower cost tiers and primary storage. Sometimes storage solutions come with automated management software to give information on how much of solid state disk you need to optimize your application. Virtualization of your storage also can go a long way in helping with storage consolidation by providing better storage distribution among servers and reducing hardware requirements. Vendors that provide server virtualization often offer tools to improve the management of storage as well.

The latest buzz in the storage world is data de-duplication also called single instance storage. It significantly reduces the amount of physical disk space needed for backup and other archival functions by configuring data processing and storage functions in a way that removes redundant files, bytes, or blocks of data. It thereby makes sure that only unique data is stored on disks and eliminates redundant files.

The other popular storage solution is thin provisioning which doles out on-demand shared storage to servers instead of allocating tens or hundreds of gigabytes in advance based on anticipated demand which may go unused while consuming a lot of energy. Provisioning helps in cost reduction as it costs less to provide power to one drive loaded to capacity than a number of them each storing a fraction of their capacity.

Please contact us at Green Rack Systems for updating your storage solutions and to reduce your energy bills in the bargain.

There is a huge initiative among companies like Google, IBM, Cisco and so many other smaller smart grid players to make the grid more modernized using cutting edge digital technology and tackle environmental issues, cost reduction and energy conservation. We employ integrated communications, sensing and measurement technologies, enhanced interfaces and control methods to optimize our electricity transmission.The common smart grid components incorporating network, wireless and IT technology consist of smart meters, data concentrators, system monitoring software, increase in transmission lines etc. They help oversee the precise location of electricity consumption and predict potential blackout areas. Consumers can also monitor more clearly where and how the electricity is being absorbed which will provide them with more leverage to effectively save on energy.

When President Barack Obama signed the $787 billion stimulus bill on Feb. 17 , 2009 , a large part of his strategy focuses on renewable alternative energy to gain energy independence but one of he largest investments, about $11 billion dollars has been dedicated to in smart grids.So it is no surprise that smart grids are being touted left, right and center by VCs and technocrats alike and promises to have a huge impact in the immediate direction of high tech and energy industry.

What does all this mean to data centers? Well according to CNetnews, this data overload due to smartgrid implementation will lead to increased demands on data center . The article cites the case of Duke Energy which has contracted with Cisco to build an “information architecture” to handle an anticipated flood of data from its smart-grid programs where it will be installing hundreds of thousands of smart meters in the next two years. Duke Chief Technology Officer anticipates that gathering data from sensors on cables, people’s appliances, and substations could add up to a millions of nodes on the network.

“Once utilities put in smart meters, their data processing and storage needs explode. Instead of sending a person to read meters once a month, information for billing or other applications can be sent back once a day, once an hour, or even every few minutes. If utilities are regulated to reduce peak time usage, their IT needs shoot up even higher. Demand response where a utility can turn down energy use at participating customer sites, requires utilities to poll information regularly from a potential large number of locations. “

Mark Weiner director of Data Center Solutions and a member of a Cisco smart-grid team is quoted in CNet news article:

“The requirements are for huge amounts of data to be involved when you have these more advanced pricing models where the goal is to mitigate power generation,” said Weiner. “The catcher’s mitt for that data is the data center.”

It is logical to conclude from the above that now more than ever there is an even more increased pressure to make your data center technology as energy efficient and cost effective as possible in anticipation of future requirements. For specialized software and hardware deployments that would help you in this effort, contact us at GreenRack Systems.

In datacenters, there is a tendency to overexert and do too much if that makes sense. After all we want to make sure that everything is working well and beyond to avoid problems during uptime. Over provisioning and over allocation often lead to overspending in terms of indulgence in excess power and hardware. Apart from virtualization, increasing storage efficiency and equipment efficiency, Power capping is a new technology that datacenters are looking into right now to address this issue. The concept is logical enough as the name suggests. Power capping sets a cap on the power a server can consume at a given time without always leaving allowances for “in case of maximum load or maximum utilization” and is uniformly applicable for all servers.

Servers come in all shapes, sizes and uses. There are many scenarios that determine their usage like customer requirements and nature of work. So they do not all run to full capacity at all times and some maybe even never as many datacenters choose to not deploy their entire arsenal at the same time. So why do we allocate all this power to them without discerning? Why do we set a standard uniform allocation irrespective of functionality? Also more times than not, the suggested power allocation referred to on hardware pieces are much more than is actually needed. Again everyone is trying to be on the safe side and being careful about not interfering with the performance level, but unwittingly wasting more resources in the bargain. That is why power capping makes a whole lot of sense. Without overstepping the power limit, it allows us to save on space and increase your server density by ascertaining that your servers do not use more than their allocated power at once.

Even though many IT managers are hesitant to experiment lest important applications get interrupted, power has proven to be such an expensive and environmentally unfriendly commodity that many datacenters have decided to take their chances on it .Companies like Intel, IBM, Dell and AMD have released power capping management software to facilitate the process. For example Intel’s Dynamic Power Node Manager Technology along with ancillary management software helps servers running on Xeon 5500 chips to vary platform power on a case by case basis without reducing the performance level through monitoring and assessing real-time data of a server platform. The ceiling on the platform power is thus adjusted to your optimal individual power budget . In case it figures that the budget cannot be met with without interfering with the performance, it will alert you on your choices. Intel also has developed a software addition to the Node Manager called Intel Datacenter Manager, designed to monitor and control power for a group of servers.

Baidu, China’s largest search company has reported success recently using Intel’s power-capping technology. In fact the companies released a study that stated that a datacenter using this technology could save up to 40 watts per system without performance impact translating into as much as 20 percent additional datacenter capacity within the same rack-level power envelope, and a potential rack-density improvement of 20 to 40 percent. Prior to investing in power capping technology, Baidu was leasing racks at a datacenter where the company sought to save money by cramming as many machines as possible into the fewest number of racks. Post Intel’s power capping tools, they were able to establish an optimal workload corresponding to maximum power reduction with least performance loss. Then they determined the approximate amount of power the servers needed to just idle. After narrowing down on the precise number of servers per rack , they successfully established a cap at the rack level bringing down the power consumption from 900W to 750W.

It is always better to understand the full implications of power capping before deciding if power capping technologies in full blast would be a viable option for your data center. A trial run before full scale operation is recommended to protect important applications. Please contact us at Green Rack Systems for further consultation.

The commercial IT vendors make a constant pitch to sell new, better and improved versions of their hardware and software every quarter. But as there is no real motivation for them to promote reuse, apart from the odd brand that exchanges older models for an upgrade or volunteers to recycle for you, we hardly get to hear much on that note.

At Green Rack Systems, we promote reuse of old equipment, as a means to not only lower the stress on landfills but also save some money in the bargain.

Eric Geier in Wi-Fi planet had some great ideas recently about the reuse of old wireless routers. He says:

Don’t throw away your old 802.11g routers just yet. The new slick-looking 802.11n routers may provide higher speeds and performance, but there are still many ways your old gear can help out, including extending your wireless coverage, improving a new network, assisting in offering public Wi-Fi, providing network authentication, or enabling secure remote or site-to-site VPN connections.

Some of these ideas consist of moving the old router around to serve a different purpose. However, most make use of the neat features that replacement firmware projects offer for the popular WRT series from Linksys and many other vendors and brands. Nevertheless, all the ideas can be quite helpful and save you a lot of money.

1) Extend coverage by using it as an AP

Though wireless routers are designed to connect to and distribute an Internet connection, they can be used just for their Wi-Fi capabilities. In other words, you can use it as an access point (AP) instead of a wireless router. If you have a small network with only a single wireless router, this can just about double your wireless coverage—and it’s essentially free.

Like a regular AP, you want to place the makeshift AP in a thoughtful spot to provide the best coverage. You want the wireless coverage boundaries of each router to overlap some. Then you have to run an Ethernet cable from the network’s router or switch all the way to the makeshift AP.

The trick to turn the router into an AP is to turn off the router’s DHCP server and to hook the Ethernet cable to one of the switch ports instead of the old router’s WAN port. For more help with this project, click here.

2) Relieve 802.11n routers from supporting 802.11g

When using 802.11n (or Draft N), it’s best to allow only 802.11n connections on the router. If 802.11g clients connect, they can slow down the newer clients. However, you can set up the old router again, just to support the old clients. You’d use the router as an AP, such as discussed above, except you could place it right next to the new router since you aren’t trying to get more coverage.

So you can better differentiate between the 802.11g and 802.11n signals, you should use different SSIDs or network names. To make sure someone loaded with a 802.11g card doesn’t accidentally connect to the new router, you can change the default wireless mode to 802.11n only.

3)Make it a repeater to take the signal further

Another way you can use an old router to extend your Wi-Fi footprint is to turn it into a repeater. Instead of having to run an Ethernet cable out to an AP, a repeater gets its network connection by listening to the airwaves and retransmitting the Wi-Fi signals between the existing wireless network and the users out of the main coverage area. This is great if you can’t or don’t want to run wires.

Though this range-extending technique doesn’t require running cables, it does require flashing your router with replacement firmware. That’s because routers don’t come with the repeater feature out of the box. If you have a supported router, you can use the DD-WRT, Tomato, or Sveasoft firmware replacement.

4) Use it as a wireless bridge

If you have computers or other network devices that need to be connected to the network but only have an Ethernet port and aren’t close enough to the router, you could convert your old router into a wireless bridge. In bridge mode, the old router would communicate with the new router via the airwaves. Any computers connected to the old router’s Ethernet ports would be just like they were wirelessly connected with the new router themselves.

Like with the repeater mode, to get this bridging capability, you’ll have to use a firmware replacement: DD-WRT, Tomato, or Sveasoft.

For more on wireless bridges, read Ask the Wi-Fi Guru, Episode XVI.

5) Offer VPN connections or connect offices together

Another feature provided by some firmware replacements is a built-in VPN server and client. This lets you set up the router for secure remote connections, so you can access files and services or secure your Wi-Fi hotspot connections. Plus if you have multiple locations, you can securely connect them via the Internet. You can find this functionality in the DD-WRT or Sveasoft firmware replacements.

6) Turn it into a hotspot

If you have a business, you could convert your old router into a hotspot gateway. Though you can simply plug in a regular wireless router to offer wireless Internet, you should implement the hotspot features. A captive portal makes users see a disclaimer or advertisements, or make payment, before getting Internet access. Plus some hotspot gateways can manage user accounts if login is required.

Both DD-WRT and Sveasoft include hotspot features. You might also want to check out the CoovaAP firmware replacement and all their free services. Sputnik offers a modified version of DD-WRT along with their paid services.

7) Make it a RADIUS server

If you want to use the enterprise mode of WPA or WPA2, but don’t have a RADIUS server to do the authentication, you may be able to convert your old router into one. If you have a supported router, you can flash it with the TinyPEAP firmware replacement.

Do your own brainstorming

We’ve discussed many ideas on how to use your old gear. Now you can review the features and documentation of the firmware projects to see if there are even more features that interest you. If all else fails, list your equipment on eBay…

Great points! For some more fresh ideas on recycling and reuse of your older datacenter hardware, please contact us at sales@GreenRackSystems.com