Archive for December 2009

Flirting with Disaster

Wednesday, December 23rd, 2009

Last week, Mark Wojtasiak at Seagate, our parent company, blogged that SMBs (and some vendors too) are flirting with disaster by not protecting their virtual servers. Although he cites research from the competition :-) , Mark makes a great point that SMBs should have a solid plan in place for recovering their critical systems (physical and virtual) in case of a disaster.

As readers of this blog know, we’ve been advocating for a while that disaster recovery is not just about the data. A disconnect seems to exist among SMBs, the media and the industry on not only the importance of DR and DR planning but also what it actually entails. Go here and here to read some of our past posts on what SMBs need to consider and can do in order to protect both their data AND servers from disaster, whether on premise, in the cloud or both. Happy holidays!

Posted by John Sun

Share

Experts Corner: Q&A with Tony Asaro at ContemplatingIT

Friday, December 18th, 2009

Hot on the heels of our last installment comes the next in our Experts Corner Q&A series with industry thought leaders discussing the world of storage and the future of the industry.  We recently caught up with Tony Asaro of Contemplating IT to ask a few questions on where he sees the industry heading, especially regarding the cloud, managing data and the future of storage. Tony has been in the IT industry for over 23 yeas as a system engineer, product manager, business development and marketing executive, entrepreneur, and as an industry analyst and consultant. Tony has focused heavily on storage, data management and virtualization technologies and products.

i365: In a recent blog post, you discuss the hype around Information Lifecycle Management a few years back and how that hype has declined. You also point out, however, that many of ILM’s ideas–matching data to the right storage environment in order to balance cost, performance, and security–have been realized to some extent.  How does cloud storage fit into this as a new way by which to store data?

TA: It is important to define cloud storage. There are two kinds of clouds – public and private.  Many companies are considering how to make their IT environments more of a utility by creating private clouds. Whereas, public clouds are infrastructure resources – compute and storage – that are accessible over the Internet.

Using a lower tier of storage is already a part of ILM but the technology has not quite lived to the promise of the concept.  This is in great part because with the amount of data being created there is no way to manually assess the use and value of data.  Therefore, the storage systems themselves need to be smart enough to determine – based on policy – where to place data.  However, the vast majority of storage systems do not have this capability.  There are exceptions and I believe that 2010 will be the year of storage-based ILM or what I call intelligent tiered storage.

Private and public cloud storage can play a role as a lower tier of storage.  But we must have the technologies to determine what needs to be moved, the ability to move it transparently, and the cost structures have to be simple to understand and justify.

i365: Following up on that, do you see the hype around “cloud computing” going the same way of ILM and only having a moderate effect on businesses? Or is cloud computing/cloud storage a realization of something bigger?

TA: Before I answer that question I think it is important to note my answer to your first question. I believe that storage-based ILM is going to be a big deal in 2010 and beyond. It is often the case with technology that it takes time for it catch up with the hype. And I believe the impact will be tremendous, saving enormous amounts of money for customers.

The cloud hype eclipses ILM but so does the opportunity. According to IDC the cloud market is already tens of billions of dollars. But they define IT clouds as online application services as well as services such as CPU and storage. So it all comes down to your definition and therein lies the confusion. We are all saying the same words but in many cases it means different things to different people.

Some IT professionals consider IT clouds to be their virtual data center. If that is the case then we have IT clouds all over the place. And it will be the evolution of the virtual data center that will be important. The more we virtualize physical infrastructure, the greater levels of optimization, utilization, agility and flexibility. Additionally, we need to be mindful of managing, analyzing and being able to get information on our virtual data centers as well. It is also important to have multi-tenancy to create true IT utilities – creating virtual domains so physical infrastructure can be shared without loss of performance, reliability or security.

Public IT clouds are already enabling new businesses. Amazon S3 has over a billion objects stored on their infrastructure and that number is growing. I also see public IT clouds as a complement to customer data centers and their private IT clouds over time.

i365: Your post “Your Data Needs Diet and Exercise!” discusses the pressing need for the storage ecosystem – both vendors and customers – to work on new ways to make the exponential increase in worldwide data manageable. Are there any major innovations you see happening in the next few years that will make this happen?

TA: It is pretty amazing how much data we store compared to how much we actually use. We talk about the growth of data but really what we are talking about is the growth of unusable data. But we keep all of our data because it is relatively inexpensive for us to do so and the risk of losing something we may need for business or compliance offers enough risk to justify it. Additionally, since data is growing at such a rapid rate and companies are dealing with 100s of TBs and PBs, there is no way to classify data. Therefore we just keep storing lots of lots of data that we will never use again. I contend that if we are going to keep storing all this data we might as well make some use of it.

Okay – to answer your question there are a few things that need to happen. We already have storage technologies that make better use of physical storage infrastructure such as thin provisioning and data deduplication. As I mentioned above I believe that a number of storage systems will provide granular and automated tiering. That is the optimization side.

In order to make our data more useful we need to put structure around unstructured content. I believe that one of the biggest advancements will be to have technologies and tools that integrate unstructured content into databases and enterprise content management systems.

i365: Given your extensive background and expertise in the storage world, it would be great to hear your perspective on what big trends you see on the horizon and how you see them affecting current trends and existing infrastructures.

TA: I mentioned intelligent tiered storage. I think this will be a big deal in 2010 and it can save a significant amount of costs in the data center.  For the larger shops it can result in millions saved.

Disk-to-disk backup with data dedupe will graduate to the mainstream and will continue on its path to being pervasive.

Greater integration between the unstructured and structured content tools and applications.

FCoE on its surface is a good thing – consolidating the network, potentially reducing cable spaghetti, reducing infrastructure costs, etc. I think the bigger implications of a converged network is creating correlation between protocols. In other words, with data storage giving us the ability to have object and file-level awareness while maintaining the performance of SAN. With FCoE – all storage protocols will share the same infrastructure and the realization of this can be achieved cost effectively.  This will take time and won’t happen next year but we can begin to move in that direction.

Share

Experts Corner: Q&A with Greg Schulz at StorageIO

Friday, December 11th, 2009

This is the second installment of our Experts Corner Q&A series with industry thought leaders to glean insight into the world of storage and the future of the industry.  This past week we caught up with Greg Schulz at StorageIO to ask a few questions on where he sees the industry heading especially regarding the cloud, green IT and the future of storage. Leveraging almost 30 years in IT, Greg is the founder of the independent IT advisory and consultancy firm Server and StorageIO, and has worked for IT organizations involved with applications, server, and storage networking. His practice leverages his experience including public and private clouds, green it, infrastructure optimization and efficiency, infrastructure resource management including performance capacity planning, BC/DR, security and backup along with virtualization among others. Greg is also author of the books “The Green and Virtual Data Center” and “Resilient Storage Networks – Designing Flexible Scalable Data Infrastructures”.

i365: You frequently discuss your middle-of-the-road stance when it comes to “the cloud” on your blog.  Aside from the current hype surrounding cloud computing, is there any other reason you believe that organizations shouldn’t take advantage of some of the boons cloud storage has to offer?

GS: I tend to take a practical or applied approach having been both customer and vendor, thus with clouds, I see the hype, however as a user of cloud and other managed services for some things, I see how they can compliment, perhaps even for some, serve as a replacement. However, with all that being said, look before you leap, that is, cloud based data protection is not a replacement for on-site data protection, rather a compliment.

Also, a good dose of common-sense data protection (CDP) should also be applied. What I mean by this is that any information that is important should be protected, if it needs to be protected, there should be a copy both on and off-site. For example if you are doing local backups today, get a copy off site using a combination of removable media and/or electronic copies to a cloud or backup managed service provider. What this means is that I do not rely completely on clouds, or, on traditional on-site with removable media backups, rather a combination. Call it belt and suspenders or what you like; however it works as I have had to recover data using both approaches!

i365: Does that mean you practice what you preach than when it comes to backup and clouds?

GS: Absolutely! Perhaps it’s the vestige of having been a customer working in various IT organizations responsible for backup, data protection, BC/DR or having been the vendor, however I do local backups to a combination of disk to disk as well as disk to removable media that goes off-site along with regular copies of more frequently changing data to a cloud managed service provider. I have even experimented with doing backups while on an airplane using services such as gogo wifi to a managed service provider; I guess that has to be a cloud backup story or what ;) … FTC disclosure: I’m paying for, and using another service provider’s cloud or managed solution, however I do use Seagate disk drives that I have purchased for different functions including local backups.

i365: As you’ve mentioned in your blog, cloud computing is an excellent way to complement existing IT structures; while it has been adopted by both large and small businesses, many companies still have fears of accepting it as the main way they conduct business.  Do you see any time in the near future where organizations will rely on SaaS or other cloud storage services to conduct the majority of their work? What are the benefits for adopting a SaaS or cloud storage strategy?

GS: There is a lot of cloud confusion out in the market place, thus customers are not quite sure what to make of it all. Likewise there has been some well publicized cloud related incidents, or, situations where there is some perhaps un-due guilt by association for clouds involving service disruptions or loss of data accessibility, perhaps even actual data loss.

Given all of the hype around clouds, it should be no surprise that not if, rather when something goes wrong, there will be a lightning rod effect. As the storm clouds clear and organizations understand where, what, when and why to use cloud, SaaS, XaaS, backup as a service, managed as a service or whatever you wish to call it in a complimentary fashion, they will see the value.

Don’t be scared of clouds, however look before you leap! Likewise take hype with a grain of salt, perhaps skepticism vs. cynicism, however too often those get confused as being one and the same.

So the big benefit and value proposition and opportunity for cloud or backup service providers is to articulate on one hand if you are not doing anything, start doing backups along with what are some best practices. On the other hand, for those already doing backups, show how cloud or managed service backups are complimentary. Backup providers also need to articulate more of their value adds including service level agreements, data integrity, availability and accessibility along with security stories as opposed to simply offering a free or low cost service.

i365: Recently you presented at Storage Decisions in New York on the subject of “The Other Green” to a large group of IT Professionals in New York.  The presentation involved different ways in which organizations can slim/consolidate their data profiles while still maintaining fast data response. Amongst the variety of tiered storage options (SSD, SAS, tape, cloud storage) and data management strategies (archive, compress, dedupe) discussed, did you find that people showed the most interest in a certain one?  What were the different reactions to the options and strategies discussed?

GS: There continues to be a lot of confusion in the industry about Green IT, storage efficiency and optimization. The Green Gap continues which is the perception that Green is only about recycling or reducing your carbon footprint, something that many organizations would like to do, however given economic realities today and lack of real regulations, focus is elsewhere. Yet when I ask IT pros if they have to address power, cooling or floor-space footprint constraints, budgets, storing and processing more information in a shorter amount of time while boosting productivity, the majorities chime in as that is what they need to do.

Customers don’t realize that by boosting performance using less energy that the result is energy efficiency with a benefit of enhancing business economics as well as helping the environment. That is an example of the green gap, green IT is and should be focused more on boosting productivity, doing more work, processing more information, storingmore data in a smaller footprint for longer periods of time at a lower cost without negatively impacting customer or service experience.

Likewise storage optimization is thought to be about spinning disk drives down to avoid power, compressing and deduping or thin provisioning, all of which are valid for addressing space or capacity concerns. However there is also the focus of addressing time or performance to get more work done in an efficient manner.

What’s fun to see is when I talk with IT pros, the customers that is, the messages and themes resonate around boosting performance in addition to capacity optimization, however it’s a different play or scenario that many vendors are not yet familiar with or ready to tell the story, so what gets told is what they know or have read about.

With time, watch for more focus around performance and productivity as forms of efficiency, optimization, Green IT as part of the next wave which includes life beyond consolation!

i365: Are there any developments you see on the tech horizon that could become the next major IT hype machine?  If so, how might these affect current/emerging technologies?

GS: There will be several, some of which will be more hype including technologies that seem to get re-hyped every year as the next big thing such as T10 Object Storage Device (OSD), holographic storage to name a few. Then there will be more technologies added to the zombie list, you know, those that are declared dead yet continue to be bought and used by customers, not to mention enhanced and sold by vendors as they are productive and profitable. Examples include tape, fibre channel, disk drives, mainframes, printers, windows, RAID and even block storage among others.

Some technologies that I think have legs that are still very early in their maturity cycle however that have very bright futures include among others PCI I/O virtualization (IOV), Fibre Channel over Ethernet (FCoE), 6Gb SAS for attaching clusters of servers to shared external storage systems, SAS disk drivers, flash and ram based hybrid SSD, cloud based technologies, virtualization expanding focus from just consolidation to enabling scale up, scale out along with agility and ease of management.

Dedupe enabled storage continues to gain traction and I think we have only scratched the surface in terms of actual market opportunity, same with virtual tape libraries for bridging the old with the new. What I mean by that is virtual tape libraries, regardless of presenting block tape or NFS file system interfaces are great for bridging to existing backup process and procedures (the old) with the new including on-site disk as well as off-site storage at a cloud or managed service provider.

Another area to keep an eye is overall end-to-end (E2E) cross technology domain (e.g. servers, storage, networking, hardware, software, local and remote services) infrastructure resource management (IRM) techniques including server/storage resource analysis (SRA) tools. Where resource management and reporting tools are convenient for showing you what you have, as environments continue to scale in terms of size or complexity, tools that can provide insight and awareness across different technology domains become more important, not just for reporting, but also for connecting the dots and finding where the real issues and opportunities are.

Share

Online Data Protection: What Are You Paying For?

Thursday, December 3rd, 2009

Cloud Services represent a significant shift in the economics of technology. Small to mid-size businesses can enjoy previously unattainable economies of scale by resourcing compute, storage, bandwidth and power from a cloud provider. If all these resources are commodities, why is it that cloud vendors charge wildly varied prices for their services? It’s the delivery of the cloud service that matters, especially in the realm of online data protection.

Start with You

To assess the value of an online backup service start by answering two deceptively simple questions:

  1. What is your current situation?
  2. What needs to change?

Most businesses completely skip this assessment. Those that are more thorough in their data protection selection process still often struggle. They run into knowledge barriers or get bogged down in details. To understand your current situation, gauge the value of your data and your in-house skill set. Questions to help with this process are:

  1. Which data, applications, and servers will, if lost, impact your business functions?
  2. What is the technology hosting this information (e.g. SQL Server 2008 atop Windows 2008 on an HP server with 2TB of internal storage)?
  3. In the event of a loss, how quickly could you get the data and functionality running again?
  4. Who would be involved in the recovery process?
  5. Where would you recover should you lose your data? Server? Site?

Answering the above five questions is easy. The next step begins the battle. What needs to change? What business functions are material? How fast is fast enough? What technology will the company adopt in 12 months?

These are questions about which every stakeholder in your company will have a different opinion. There is no one right answer. Understand the tradeoffs on each of these dimensions so that you can make the proper value assessment when selecting a backup solution.

Understand the Value of a Data Protection Service

With understanding of the business’ situation and priorities, compare your requirements to features offered by a data protection package. A backup solution’s price is composed of four main elements:

  • Infrastructure
  • Software
  • Management
  • Support

Infrastructure: Hardware, storage, rack space, bandwidth, and power are standard components offered in a cloud service. However, a broad range of choices can be made when provisioning these basic elements. iSCSI or Fibre Channel connectivity? This choice is a price-performance tradeoff that can severely impact backup windows and restore times. Backup and restore times will be similarly affected by available cloud bandwidth. What tier is the datacenter? Finally,  is it important that your service provider undergo regular audits to achieve SAS 70 certification?

Software: What data protection capabilities are offered by the solution? Some services protect only Windows servers and desktops. Others can cover more heterogeneous environments that include AIX, IBMi, Linux, and/or virtualization platforms.

What level on the application stack requires protection? Should you need to replace hardware, you may spend hours and days reinstalling the OS, applying patches, reinstalling drivers and applications, and reconfiguring the machine before you can recover the data. Bare metal recovery software automates this work for you.  Moving to the top of the stack, Oracle databases, Exchange, SQL Server, SharePoint all require custom protection to guarantee optimized backup performance as well as full application recoverability. When evaluating various solutions, understand what it will take to recover your servers, data and applications.

How quickly will you need to recover? Internet recoveries are limited by the internet connection speeds as well as network latencies. Does the solution offer the ability to physically ship data to your site? Is there an option to store a cache of the data onsite for quick, LAN-speed recoveries?

Management: Control of the data protection solution from a single, centralized location minimizes the effort required to reconfigure backup jobs and schedules. Backup storage reporting helps manage your monthly storage bill. Automated aging of the backup data rounds out external-facing management functions.

What management is happening behind the scenes? Data validation, optimization of the backup pools and server health checks are tasks completed on a daily basis. Less frequently new storage is provisioned, servers are replaced and software is updated. The hours devoted to these processes ensure compliance with the SLAs specified for Tier III and Tier IV datacenters.

Support: Support spans the lifecycle of your data protection solution: from selection to deployment to regular operations. What level of education will your business need to choose a solution, and where can you obtain that guidance? Who will be responsible for deployment and installation? Is training available with the package? What is the average turnaround time on your operational questions and issues? In the event of a disaster, how quickly can you get a live person on the phone to ship a copy of your data? Make sure to obtain the correct balance of support with your in-house expertise.

Remain Employed

Selecting a data protection solution is not easy. Today’s economic climate dictates that the cheaper the solution the better. However, before diving into a backup selection process, understand your company’s situation and risk tolerance. Work with your vendor to understand the value offered in the areas of Infrastructure, Software, Management and Support. And don’t back yourself into a corner with your selection. All too often when we ask prospects what they would do should their application / server / site goes down, their answer is “Go look for another job.”

Posted by Karen Jaworski

Share