by admin on April 2, 2010
Thanks for coming back! While here, you may want our Free Whitepaper.
We’ve talked here about the inevitability of storage growth and concluded that it’s simply unmanageable without a comprehensive view of the entire storage infrastructure. And, as storage becomes a concern for a broader range of companies – and not just the biggest – more and more folks are going to look for ways to optimize the utilization of their storage assets and stop using their most expensive storage for low value data.
Sound familiar? Wondering what to do about it? If you haven’t looked at tiering your storage and the benefits that can bring, it’s probably worth a look. If you’re not that familiar with storage tiering, it involves not only data discovery and classification, but also typically requires some means of performing ongoing analysis across a heterogeneous storage infrastructure.
First, the basics… tiered storage means putting only high value, mission critical data on the most expensive tier one storage media such as Fibre Channel while using lower cost storage – SAS, SATA or tape – for lower value data. The expense of the storage should correlate to the value of the data. Moving data to more appropriate storage media can improve total cost of ownership. It will likely also improve storage performance. If everything is stored on one storage tier, there is only one resource to meet user demand for data access. When spread across multiple storage media, the burden of user demand is spread across the resources so users may even experience improved performance.
The first step is determining what makes data valuable for your organization. Perhaps tier one data is anything that’s been accessed in the past week or is otherwise determined important or mission-critical. Tier two data may be financial files or anything less frequently accessed. And there may even be a third tier of data including rarely used or unclassified data. Once you have defined the value of data, it’s still far too difficult and time-consuming to weed through each and every piece of data individually to determine its value to the organization. But, in order to get the true benefit of tiered storage, you have to have some way of determining how much low value data is taking up precious and expensive space on tier one storage media and move it off to lower cost storage.
And this is where a lot of point storage tools and solutions fall short. If you only have access to tools that discreetly manage each different type of media, there is too much room for error. You could miss large chunks of data, or run into problems trying to reconcile the various monitoring platforms reports into a single view of the environment. A solution that can perform analysis across the entire heterogeneous storage environment allows you to capture all of the data you’re looking for, see what type of storage it’s on, and move it according to your data classification policies.
With a single management view for a heterogeneous environment and some relatively simple file analysis, companies can begin to effect positive change in their organizations with tiered storage.
Our philosophy has always been that offering folks the ability to monitor and manage everything in their environment from a single pane of glass is the most cost-effective and productive way to maximize your investment. This is true of SolarWinds’ network management offerings, and is one of the things that made Tek-Tools a great addition to the SolarWinds family.
Follow us @solarwinds_inc
by admin on March 17, 2010
[Ed. NOTE: Don't forget to follow us at @solarwinds_inc and @sw_headgeek]
The growth predictions for the storage market never cease to amaze. A recent G*rtner report [“Controlling the Cost of Storage Growth Requires Storage Modernization” (March 15, 2010) penned by Bob Passmore] predicted that storage growth is expected to return to the 60% to 70% range as the economy improves. And yet, businesses still have not figured out how to effectively manage this? Is there really a business value in hanging on to all of this data?
Not many would argue that it’s all valuable. Most would say that it’s just too difficult and time consuming to figure out what’s valuable and what’s not, so it all just gets boxed up and thrown in a climate controlled storage unit (so to speak). Years later, the unit gets opened up and no one can remember why they’ve been paying all this money to store the stuff.
Of course, there is a lot of data storage that is absolutely necessary. It’s needed for day-to-day business operations or to achieve compliance with any number of legal and business data archival requirements. But, a lot of storage capacity is wasted on volumes upon volumes of non-critical data. It’s all the ad hoc content that users are constantly creating. It has become so easy to generate content – pictures, videos, presentations – and share it with others that multiple copies of the same or very similar files are often being stored within an enterprise. Based on conversations we have had with our customers, it wouldn’t be out of line to estimate that non-critical data like this can take up 50 percent or more of an enterprise’s storage capacity.
Some may ask, “Shouldn’t data retention policies take care of this as a way to be sure only critical business data is stored after a certain period of time and non-critical data is deleted?” In theory, yes, and to be fair, most enterprises do have data retention policies in place, but too often they have no means to really enforce them. Plus, they typically only apply to critical business data so non-critical data is left to take up valuable storage space.
Others may look to data deduplication and thin provisioning technologies to help. And they can. Data deduplication can significantly improve storage utilization by indexing and deleting duplicate copies of data, but it simply won’t be able to get rid of everything that’s unnecessary. And thin provisioning can help to optimize utilization of available storage. But, getting a true handle on storage growth means understanding what’s in the environment in the first place.
That’s the real problem – especially with heterogeneous storage environments, but it is step one and it cannot be skipped. From an admin’s point of view, if they can’t see what they have in the environment, how do they manage it? Without a comprehensive view of the entire storage infrastructure - not just individual storage assets - including visibility through optimization technologies, admins are hard pressed to manage storage growth efficiently. Once they have a complete picture of what’s in the environment and can see how storage is growing and how users are consuming data, then they can get down to the business of really managing the storage infrastructure effectively.
Find out more at SolarWinds and Tek-Tools.
by sw_headgeek on March 4, 2010
[Ed NOTE: Today's post comes courtesy of SolarWinds' Head Geek, Josh Stephens. He offers his special blend of insight and perspective on day to day IT operations management over on the Geek Speak Blog. Follow Josh here --> @sw_headgeek.]
Today, let’s talk about the need for insight into the entire picture in order to optimize performance. Whether you’re a small shop with a simple IT infrastructure or a large enterprise with technology and equipment taking up big-time real estate in your buildings, performance is critical, especially application performance. At the end of the day, it all comes down to application performance.
Users expect near instant and constant access to the applications they need to do their jobs whether they’re working on site or (increasingly) off site. Generally speaking, they see no reason that their applications and data shouldn’t just be there when they log in, regardless of location. And why shouldn’t they expect this? Companies have an obligation to provide the resources and tools necessary to get the job done, don’t they?
Herein lies the rub for IT departments everywhere – how to ensure application performance with zero to low threshold for anything less than 100 percent up-time. An application performance issue is still often identified when the users start calling the help desk complaining that an application is “slow.” “Slow” meaning that the speed of the retrieval or manipulation of data within the application is taking more time than the user is willing to tolerate. (Granted, totally subjective… some users are more patient than others.)
Nonetheless, the problem for IT teams is that a performance issue may have little to do with the application itself, so where to begin to start to figure this out? Application performance can be affected by the network, server, storage, as well as the application itself. The only way to effectively diagnose the issue is to have visibility across the infrastructure.
To get to the bottom of a performance issue, you have to understand the interdependency of multiple infrastructure components – servers, storage, backup, applications and the network – and the effect each can have on application performance. If you don’t know what’s connected to what, you’ll be hard pressed to find the problem or at least to find it quickly. Add to this the layer of abstraction over everything in today’s increasingly virtualized infrastructures and you have a real challenge to address. A comprehensive view of the interdependency of the infrastructure components allows IT admins to pinpoint the true root cause of a performance issue and solve the problem quickly.
And, real-time data is just the tip of the iceberg… it certainly helps identify right away if the issue is indeed with the application. But, IT admins need to be able to see both historic and real-time application performance data. Historic data allows for analysis than can arm IT teams with the data they need to assess and make decisions around trending. Plus, the ability to conduct performance analysis of multiple components simultaneously is great for load-balancing and identifying bottlenecks to avoid issues in the future.
So the next time users are complaining about how slow their email is, before you blame the application… or the server… or the storage… or the network… better yet, before the next time, get performance visibility with a solution that can monitor all the components of your infrastructure.
Flame on…
Josh
Follow me on Twitter
Learn more at www.solarwinds.com