Posted by Crystal Nichols on Mon, Jul 26, 2010 @ 08:56 AM
As any data center manager or CIO can tell you, data center power costs have become a huge concern. Being able to measure the efficiency of power usage at the data center and being able to make improvements can reduce energy costs and also create an environment in which data center equipment operates at a higher level of efficiency. The good news for data center managements is that there’s a new way to measure power on the way.
A number of organizations, spearheaded by The Green Grid, has created a method for data centers to be able to measure the efficiency of their power expenditures. The standard, known as Power Usage Effectiveness (PUE), is designed to help bring some consistency to the way that data centers measure power usage, and to make it so that different data centers can compare their numbers.
The standard creates the PUE measurement by comparing the actual energy used by a data center to the amount of energy that actually makes it to the IT equipment. This demonstrates how much power is consumed by ancillary systems, such as cooling systems.
There’s been talk of using PUE in the data center for some time, but up until now there hasn’t been a way to compare it in a standard way. This method for calculating PUE was developed by The Green Grid along with the U.S. Department of Energy, the Environmental Protection Agency's Energy Star Program, the Uptime Institute, the U.S. Building Council and more.
For now, the group has only released recommendations for dedicated data centers. The plan is to provide another set of recommendations for use in data centers that are not part of a dedicated facility, but rather are one part of a mixed-use facility.
This new standard will help to create a common understanding in the industry that should foster dialogue and will, in the long run, create more efficient data centers. It utilizes four different ways of tracking PUE so as to include as many types of data centers as possible. The lowest level of measurement uses the utility meter to take snapshots over 12 12 months, while the most accurate type of PUE measurement will measure at the point where devices connect to the electrical distribution system.
Green Networking in the Data Center
Download this white paper from HP for more information about:
- The challenge of a complex data center
- Green data center best practices
- Power and cooling utilization
- Building efficient infrastructures
Posted by Crystal Nichols on Tue, Jul 20, 2010 @ 06:42 AM
It’s easy to get stuck in an infinite retention trap. Absent a unified storage management and retention policy and strategy, you wind up erring on the side of caution and saving everything forever.
This is, of course, the retention policy that takes the least amount of effort, at least in terms of design. It’s also the least efficient. When you keep all of your information forever, you’re going to have higher costs in the end. You need to maintain every application tied to all of your retained information, and eventually you’re going to find your data center bursting at the seams.
The folks at HP are familiar with the infinite retention trap. Accordingly, they offer a handful of ideas about how to get started in creating an information retirement policy:
1. Know what your data needs are. You need to analyze how much data you have today, as well as how fast your data needs are growing. At some point down the road, you’re going to hit your limit. That limit might be one of physical space in the data center. It might be a limitation in manpower to manage your burgeoning infrastructure. In most cases, that limit is going to be budgetary.
2. Know what kind of information you have. Not all information is created equal. You need to bring in experts from your various business units to categorize every piece of data. You need to classify data as business records or non-records, and then proceed to create a specific policy of retention and retirement for each category.
3. Identify your regulator obligations. Once you have your categories of information, you need to start thinking about how regulatory requirements affect each of those categories. Regulations will vary greatly depending on your industry, as well as the type of data you’re talking about and how you use that data.
4. Choose your vendor carefully. You need to choose the right solution for your needs. You want a solution that can implement your retirement and retention policies through the entire life of the record. You need a solution that will support multiple types of data, including email and business applications. Whenever possible, your solution should include an open data format so you can access your data through a variety of means. Finally, you need to choose a vendor that’s going to be around through the life of your data, so they need to be one that’s stable and that isn’t going to shut down any time soon.
There are many vendors that can help you implement these kindsof storage retention policies, and HP is at the top of the list when it comes to providing robust solutions in an affordable way.
Posted by Crystal Nichols on Thu, Jul 15, 2010 @ 08:44 AM
IBM POWER7 promises to be one of the most robust, flexible and efficient technologies offered by IBM to the enterprise to date. If your organization is thinking about making the move from a previous IBM Power Systems configuration to POWER7, there are some specific things you’ll want to watch out for in order to keep things moving along the way that they should.
Power5 Upgrades
As it stands today, any upgrade for your Power Systems that intend to preserve the same serial number when going to POWER7 will have to come from a POWER6 server. What this means is that, if you’re on POWER5 or older technology, you’ll have to add the interim step of upgrading to POWER6 prior to upgrading to Power7, if you intend to be able to keep your serial number.
I/O Drawer Support
POWER7 systems will be able to support the current POWER6 system 12X I/O drawers. Older drawers, such as the slower RIO/HSL-attached I/O drawers, won’t be supported on POWER7. What that means for you is that you should consider replacing your old slower RIO/HSL-attached I/O drawers with drawers featuring newer technology.
SCSI Disk Drive Support
When POWER6 was released, IBM announced that this would be the final line of servers to support SCSI drives at 36GM or smaller. In addition, it was the last to support 10k rpm SCSI drives. If you still have smaller SCSI drives or slower SCSI drives, you’ll want to consider replacing them with newer, supported drives prior to your POWER7 adoption.
QIC Tape
As with SCSI, POWER6 was the last IBM series to support usage of Quarter Inch Cartridge (QIC) tape (also known as SLR) drives. Here again, if you want to go POWER7 you’re going to need to move to newer media, drives or technology prior to your upgrade.
PCI Adapters
POWER6 also signals the end of the line to support the IBM I IOP and IOP-based adapters. These adapters can be replaced with IOPless, also known as “Smart IOA,” options. Moving off IOPs will help make your implementation of POWER7 go much more smoothly.
Some of the problem comes with device support for IOP-based tape libraries, for example. Other problems come with the need for an IOP from SDLC or X.25 on WAN and LAN adapters. Check to see if your current hardware requires an IOP before you migrate to POWER7. If it does, that hardware will have to be replaced, as well.
Posted by Crystal Nichols on Tue, Jul 06, 2010 @ 08:49 AM
While there are signs that the economic downturn may be finally beginning to turn around, the fact is that IT departments are, for the foreseeable future, going to be asked to offer more services with higher performance for less money. As technology advances, some costs naturally come down, and so there are some built-in ways that the market is going to help IT departments manage. Still, you’ve got to cut some corners here and there, and one of the best ways to do that is in the area of your maintenance contracts.
Support contracts themselves can be a cost savings measure. If you have an application or a particular solution and you don’t have any in-house expertise, you have a few choices. You can repurpose existing personnel to manage that technology. This creates a drain on your personnel, however. You can bring in new personnel to support the technology, but in many cases those positions may be very temporary or not need a full-time person to administer them.
So, you buy a maintenance contract. You save a bunch of money when compared to hiring someone new, and you get expert support from folks who deal with that particular technology day in and day out.
But, even these contracts add up. And, depending on exactly the extent of your inventory of solutions and contracts, it can be hard to keep your arms wrapped around your inventory. That’s where it can become worthwhile to begin consolidating your maintenance contracts. Consolidating your maintenance contracts with a single, reputable support vendor provides you with a number of immediate benefits. Your contracts are contained within a single coterminous agreement. You have a single point of contact when it comes to external support. There are fewer invoices and fewer agreements, therefore there are fewer opportunities for error. Not only that, you now have a support resource that is not only familiar with the specific technology that’s being supported, but rather that has a handle on your entire IT structure and strategy.
You’ll also realize immediate cost savings, as a consolidated maintenance contract will be less expensive than several separate contracts. You can likely negotiate even better pricing when you make a multi-year agreement, or when you add additional systems and technologies to the support contract.
If your maintenance contract situation is a mess, talk to your favorite reliable IT vendor today about cleaning things up with consolidated maintenance contracts.
Consolidate your Maintenance Contracts with the Intelligent Help Desk
The Intelligent Help DeskTM is a telephone and web accessible support service center available for the exclusive use of Unitiv clients—24 hours a day, 7 days a week. The Intelligent Help DeskTM provides a single point of contact for technical support and trouble resolution across multiple network computing architectures and information system manufacturers.
Click here for more information.