Asigra Inc.

Storage Discovery Needed Before Backup Redesign Can Begin

2009-03-24T10:00:00Z

One of the most vexing problems in enterprise data centers today is the lack of information that those in charge of data centers have about the infrastructure that they manage. When I used to work for a Fortune 500 company it was like pulling teeth to try to get the base line information that I needed even in order to understand how to manage the infrastructure and this was before the advent of server virtualization. These problems certainly have not decreased and, if anything, organizations now have less time, less money and probably less people but a greater need than ever to understand their infrastructure so they can manage it.

The need for information becomes especially critical as more organizations make backup redesign their top priority. Yet before they can properly redesign their backup infrastructure, information about sizing their production infrastructure is needed, such as:

The total amount of data on the network
The breakdown of file types
How frequently the data is accessed
How much of the data is redundant
Data growth and change rates

Of course, finding a tool that can quickly and efficiently produce this information can be as or more difficult as redesigning the backup infrastructure itself. Then even if organizations do find a tool that can produce this information, it often requires that they first deploy agents on each server in their environment.

This approach may defeat the whole purpose of deploying the tool in the first place since the IT staff first has to do what the tool should do: analyze the data on the network to estimate the size of backup solution So due to the length of time it takes to deploy the software and the difficulty of the task, organizations should not skip this critical sizing step in order to have an optimized data storage redesign and cost-effective backup procedures.

Asigra Televaulting offers a LAN Storage Discovery feature that is useful both before and after an organization begins its backup redesign. Before an organization makes any changes to its infrastructure, this module can first be used to analyze an organization's entire concentration of data on the LAN.

Using an agentless approach, it sweeps the organization's network multiple times and generates reports that help organizations identify possible storage inefficiencies, such as stale or redundant data on these servers. Using these reports, organizations can quantify the amount of data on their network and estimate the size of the backup solution that they need.

The LAN Storage Discovery tool remains valuable even after the initial discovery is complete. Once a new backup solution is put into place, it can continue to help an organization improve their backup infrastructure. It can identify static and/or duplicate files that once backed up, no longer need to be backed up which contributes to reducing backup windows and backup data stores. It also can keep track of the number of files that change over time and how frequently they change so organizations can plan for data growth on both production storage and storage used for their backup infrastructure.

Data centers are often stuck between a rock and a hard place by being forced to protect an organization's more precious assets, its data and information, without having the tools it needs to report on and manage the infrastructure that stores this data. Asigra Televaulting's LAN Storage Discovery changes this dynamic. With its agentless design Televaulting can quickly and easily scan the corporate network to provide the preliminary information that organizations need to quantify the size of their production environment, size their backup infrastructure and then continue to keep it optimally tuned and sized for future growth.

A New Option for Enterprise Data Protection

2009-02-05T11:00:00Z

Enterprise data protection software is experiencing a fundamental shift in terms of what organizations expect it to deliver and the amount of distributed structured and unstructured data that it needs to protect. As recently as a few years ago, the expectations of enterprise organizations were relatively modest - support for most major operating systems, integration with major applications (MS Exchange, Oracle, etc.) and tape library support - as compared to today's standards. While some of those requirements still hold true today, more has changed than has stayed the same. This is putting a great deal of pressure on data protection products to swiftly evolve.

Here are just some of the new demands that organizations now have for their enterprise data protection requirements that did not exist even a few years ago.

The management of backup data on disk. Most data protection software products support the backup of data to disk but that is only part of the challenge today. To store data to disk and keep its costs under control, compressing and deduplicating the data are becoming prerequisites. Other considerations with storing backup data to disk include managing different tiers of disk, replicating data between sites and, in performance intensive environments, using high-performance disk during peak backup and recovery windows.
Streamlined backup and recovery processes. Organizations want to protect more data, spend less on data protection, use less people to manage it and backup and recover their data in less time. These new requirements challenge the traditional model of using dedicated backup administrators, installing agents on all of the servers and continually managing the hardware and software that make up the backup infrastructure.
Index the backup data. New federal and state regulations are putting new requirements on organizations to access, index and retrieve their data regardless of where it resides, including backup repositories.

These are just a few examples of how expectations for enterprise data protection software have changed and, while there are obviously many more, what is interesting to note is how some data protection vendors are adapting their products to meet these new organizational demands. In many cases, data protection vendors are purchasing other products and offering these new products as loosely integrated point solutions to their existing offering. But the problem that starts to emerge with this duct-tape approach is that it can start to make the data protection software more complex. More management interfaces, more user logins and no common backend data repository or policy engine results in a less integrated product with higher TCO and a lower overall level of data protection.

In this respect, Asigra Televaulting has separated itself from other players in the enterprise data protection space. To understand how this has occurred, it is important to understand a bit of the history about how Televaulting was developed. David Farajun, Asigra's founder and CEO, lost a ton of valuable data while developing software for programmable logic controllers in mid-1980s. Asigra grew out of a disaster when a hard drive failed and David wanted to make sure that never happened to him again, so he went looking for a foolproof way of doing backups. He didn't find what he wanted - but he did find that many other people had suffered the same sort of catastrophe. At that time, the usual way of backing up to a locally attached tape device- but many people failed to backup often enough, if ever. Also, the locally backed up data on tapes had to be manually shipped off-site to a secure vault. What was needed, he thought, was a completely automated approach that backed up data to a remote location.

Asigra is the only vendor that developed its data protection around an architecture that specifically allows customers to backup files to a secure or protected site i.e. a remote data center for the purposes of creating a second copy of the customer's primary files to insure them against any loss of data. Since then, Televaulting has continued to evolve and mature under Farajun's direction such that today it is now well-positioned to deliver on the demands of today's enterprise organizations. Features such as agentless backup to disk remain while new features such as support for most enterprise operating systems and applications, backup lifecycle management, deduplication, encryption, replication and support for virtual servers are now part of the Televaulting 8.0.

This is a time when a number of enterprise data protection vendors have acquired third party software to deliver on the new data protection functions that enterprise organizations now want. But because of how Asigra was founded, Farajun's continued deep involvement in Televaulting's product development and how the requirements of enterprise data protection itself have evolved so that they play to Televaulting's strengths, enterprise organizations now find themselves with a distinctively new data protection option.

Confidently Move Into Cloud-based Backup & Recovery

2009-01-28T19:00:00Z

When I look at current IT trends and see companies like Microsoft getting into the software-as-a-service (SaaS) or cloud-based backup and recovery space, it demonstrates that most vendors see SaaS and cloud-based computing as a real marketplace on both the business or commercial sides of the house. It's evident that more end-users and businesses are increasingly looking at cloud-based backup services as one less thing they need to manage and that can be delivered as reliably and at a lower cost than what it costs them to do backup today.

Most businesses small and large have many IT needs but one that they continue to focus on as they move into a completely paperless world is data protection and, more specifically, data recovery. They know their current in-house backup and recovery processes are often less than adequate so when they ask hard questions like, "How long can I afford to be without my data?" and "What does losing that data mean to the company and the company's public reputation?", they don't like the answers. But what IT managers are surprised to learn as they look to move to a SaaS offering based on a cloud-based computing architecture for their backup and recovery services, they find there are many options from which to choose.

Yet one offering that has distinguished itself in the cloud-computing/SaaS market place is the Asigra Televaulting solution. Unlike most other offerings that are just starting out, Asigra has 20+ years in the business of outsourced data protection. Asigra already has established relationships with hundreds of managed service providers (MSPs) that each service tens if not hundreds of clients. All of these combined equal petabytes of backup data already under the management of Asigra Televaulting.

Further, by moving from an in-house backup and recovery environment into an Asigra powered cloud-based service model it reduces, if not altogether eliminates, some of the following costs associated with backup and recovery in an IT environment:

Personnel costs. Moving backup administrators and operational people to other areas adds more value to the business while it eliminates the personnel costs associated with tape media management and movement.
Infrastructure costs. By outsourcing backups and using a third-party provider, resources previously dedicated for backup can be redeployed for other purposes.. For instance, organizations can redeploy backup servers, retire, scale back or remove tape libraries and tape drives as well as the areas of the building used for physical tape storage areas
Offsite storage costs. An aspect of data protection that organizations often do not think about is how much they pay monthly and annually to store tape cartridges offsite and the ongoing costs associated with moving these tape cartridges to and from these offsite facilities.

Asigra Televaulting coupled with its existing, extensive and established network of MSPs breaks the barrier to traditional backup and recovery approaches by keeping customers' risks low. Daily I hear from end-users that they are looking for more unique and cost-effective ways of running their IT business. Moving to an outsourced cloud-based computing backup model is a great way to reduce backup and recovery costs while at the same time creating a more reliable and robust backup and recovery infrastructure.

Innovate in 2009 Using OPEX not CAPEX Funds

2009-01-12T11:00:00Z

To say that organizations are approaching 2009 with more than just a little apprehension would be an understatement. Scandals are rocking the financial markets on an almost daily basis. There is the looming threat of new legislation in 2009 which will make it more expensive to conduct business going forward. And, in the US, nearly 700,000 individuals in the private sector lost jobs in the month of December alone - Yikes! That leaves those left in organizations trying to figure out new ways to deliver the same amount of value and services with less money and people and nothing is more clearly in the sights of businesses than lowering their IT costs and keeping them under control.

By way of example, I recently received an email alert from the Pennsylvania State Employees Credit Union (PSECU). Thankfully, the credit union is on solid financial footing, added $28 million to its reserves in 2008, and has over $312 million in reserves with no exposure to subprime lending and option ARMs. But that does not mean it is about to spend a bunch of money in 2009 as it is looking more to survive than thrive in 2009. To do that, PSECU is lowering its interest rates on deposits, trimming its 2009 budget in terms of salary increases and will not expend any money on new technology unless it is required by regulators or adds to member security.

For the most part, I agree with all of the steps PSECU is taking but wary about the last one. I'm certainly not suggesting that PSECU go out and spend money on new technology like a drunken sailor. But to stop spending money on new technology, even in a tough economy, carries its own set of risks. It could put an organization behind the 8 ball when the economy picks up plus make it more difficult to lower costs should the recession continue into 2010 as some economists forecast. But part of the issue that organizations such as PSECU grapple with is how to bring in new technologies without massive capital expenditures.

In this economy, probably one of the best ways to avoid this cost is by identifying IT providers that give organizations the option to eliminate the need for upfront capital expenditures (CAPEX). Rather, they should bring in technologies where organization can use their more stable, but flexible, operating expenditures (OPEX) budget.

One way organizations can accomplish this is to look for providers who offer flexible ways of leveraging state-of-art technology either as rental licensing or as a service for which flexible operating expenses can be used by organizations that are unwilling or unable to invest large upfront capital expenditure but have a pressing business need. A logical place for any organization to start thinking in terms of OPEX budget is with data protection - especially as data is growing exponentially and upgrading the storage infrastructure and protecting the company data is not an option.

I say this because despite all of the innovation and changes the last few years, backup is still consistently listed as on the number one IT problem in organizations. Data protection that is available as a service has evolved significantly - both in the features it offers, how it is priced and its maturity - so organizations can now reasonably and confidently introduce it into their environment with all of the features they need using funds in their OPEX budget without fear of losing data or disrupting their current operating environment.

Asigra Televaulting offers its backup software as a service and licenses it on a per GB basis through its vast network of Managed Service Providers. Further, it includes many enterprise backup software features as part of its core offering, such as agentless backup, continuous data protection (CDP), compression, deduplication, disk-based backup, encryption and support for multiple applications like Microsoft Exchange and SQL Server, just to name a few so organizations immediately gain access to these features.

For enterprises that want to manage the backups in-house and do not want to leverage a service provider for any reason, Asigra offers its platform in a rental licensing format - which can be converted into a perpetual license at any point in the future. Regardless of which path an organization selects, the organization can eliminate the upfront CAPEX costs associated with implementing a new data protection product while leaving IT department some flexibility to innovate even when its purchasing hands are partially tied.

I respect every organization that is trying to keep its costs in-line and not spend money unnecessarily. But to stop innovating and bringing in new technologies in 2009 is a risky proposition short and long term. That is not to say innovation is not going to be easy in 2009 but neither is it impossible. Organizations that utilize their OPEX instead of CAPEX funds to procure data protection solutions can meet both their tactical requirements to keep their costs under control while still bringing in new technologies that address existing problems while better positioning them for the future, regardless of what that future may hold.

Dialing up the Right MSP for Online Backup Requires More than just Picking up the Phone

2009-01-06T11:00:00Z

In an earlier blog entry, I took a look at why more IT managers are choosing to rent, rather than build, their backup infrastructures by outsourcing their backups to managed service providers (MSPs). However as the number of MSPs proliferate, the decision about which MSP to dial up gets harder, not easier, since more and more VARs are jumping on the SaaS bandwagon to offer Managed Backup Services. Further, companies need to quantify their own needs and expectations as they select an MSP. Below are some examples of questions that they need to ask and answer internally and externally before making this important decision.

Internal Questions

What are the RTO and RPO requirements for my applications?
Do you have current data retention policies in place?
How long does each type of business data need to be retained for?
After a certain period of time, should the backup data be moved into an archive for longer-term retention?
What sort of backup SLA do you need to provide to your customers?
How often is your organization going to perform Disaster Recovery testing?

Questions for the MSP

What backup software have you deployed in your data center to enable the managed backup services?

This question is the key because all the features and functionality you will be leveraging when running backups and recoveries is ultimately tied to the platform they are running internally. MSPs like AmeriVault who leverage Asigra's Televaulting platform are able to offer the market leading functionality at a low cost - which is the primary reason for the success in retention and growth of their customer base.

Other questions you can ask include:

How many customers are using your backup and recovery services?
How long have you been providing this service to you customers?
Do you own or lease your data-center facility?
What type of facility will my data be living in? (SAS-70 Compliant, DC type, Network, Security, 24x7 operations support, generator backup, UPS, Fire suppression etc)
Can we take a tour of your facility?
Can we meet some of your support engineers?
Do you offer a replication option, to ensure my data exists in at least two locations?
Can I speak to some of your customers? (Especially ones in your vertical if possible)
Can you show me your metrics for uptime on your backup and recovery infrastructure down into the facility?
Can you show me some metrics on recovery times for different types of file sizes and entire systems?

There are many other questions that can be asked but you really need to base them on your specific situation. Feel out the service providers you are talking to in order to understand what specific market they are seeking to satisfy and use that information to attempt to find one that meets your specific needs. For example, you may not need a superior facility to house your data (based on your business needs, of course) while some organizations may need a Tier-4 data center. In whatever circumstance you find yourself, make sure before you only pull the trigger after you fully understand what your MSP is going to provide you.

Determining what the cost structures are for this deployment is equally important to the decision. Weigh the cost of how the MSP will charge you for its services, whether that is by TB backed up, backup jobs, restore jobs, etc. Further, do they provide multiple facility types and do your costs go down based on the robustness of the facility your data is housed in. For example, maybe your company requires a very stringent data integrity, which would probably push you in the direction of a more hardened facility.

The real positives about outsourcing your backup and recovery environment to a well respected MSP can save you substantial amounts of money and headaches - if you choose the right one. Managing your own backup and recovery infrastructure requires you manage multiple moving parts (tape Libraries, tape drives, network performance, virtual tape libraries, etc.), as well as the people, media and offsite storage facilities. So when calculating an internal versus external cost models, make sure that you are dialed into and accounting for all those aspects in mind because as you do, the rationale for going with an MSP for your backup and recovery environment can makes a lot of sense in many circumstances.

Why More IT Managers are Choosing Backup to the Cloud over Build Their Own Backup Infrastructure

2008-12-23T11:00:00Z

Offering the appropriate technology solutions to your internal business customers is a priority for any technology manager that desires to provide high levels of service at the lowest possible costs, particularly in the troubling economic times that we are living in today. However, knowing when to pull the trigger and outsource a critical IT function such as backup versus making further investments in infrastructure choice is not so cut-and-dry when your name is on the dotted line. Further, every IT manager now regularly faces the "Do I continue investing in hardware and software upgrades to support the data growth in the data center and remote locations ?", or "Should I start leveraging the backup services of a Managed Service Provider via its cloud computing offering?" conundrum.

This is a question more IT managers are asking every day if recent IDC research is any indication. In November 2008, IDC forecast that online storage was poised to take off and it counted online backup, archiving and replication services among those services as ready to roll. So as IT managers consider their options, they need to understand that while there is no right or wrong answer to the question, it is a question they need to ask and answer for themselves. The right answer to it, however, does highly depend on their business environment, IT staff expertise, customer locations, ability to expand, current infrastructure (if any) and future plans.

Let's take, for example, an IT manager for a company that is looking to outsource its backup and recovery services and is currently evaluating a Managed Service Provider (MSP) that uses Asigra Televaulting as its backup solution for its cloud computing offering. The IT manager may or may not have the needed expertise in-house to manage backups now based upon the type of IT organization that he/she has developed. However, that is just a small part of the equation.

From a facilities perspective to support backup, the IT manager probably does not have much more than a small data center, if that, with no infrastructure to speak of (UPS, FM 200, Generator, etc). Providing an adequate network infrastructure, to the backup servers and their respective clients is a real concern that needs to be taken into account also. As you can see there are many things one needs to think about prior to making a decision on in-sourcing a backup and recovery environment. With all of these parameters it's obvious why this market segment has exploded over the years.

Now, the question becomes what sort of MSP (Managed Service Provider) is right for my organization's backup and recovery needs. Choosing the proper service provider is one of the most important decisions that the IT manager needs to make prior to moving forward in this endeavor. The best idea is to make a list of your businesses needs and compare them to what the MSP can provide. Equally important, determine the state of the facilities of the MSP that will be hosting your backup data. In a forthcoming blog entry, we will take a look at some questions companies should ask of themselves and their MSP in preparing to outsource their backup infrastructure.

Arch Rivals within the Premier Football League Find Commonality in Back Office Backup

2008-12-16T11:00:00Z

Despite what happens out on the pitch, the Premier League is experiencing a small awakening amongst its clubs - and some unexpected harmony - for their IT disaster recovery solutions. The challenges, demands and expectations to deliver a robust backup and recovery solution for these clubs is just as pronounced as any other corporate datacenter. However, faced with meeting the escalating salaries of their best football players, the IT staff often comes out on the short end in these organizations.

Teams in the Premier Football League often have a high public profile but their budgets are as tight as any other organization when it comes to data protection. This leaves them with a skeleton crew that is unable to meet the RTOs and RPOs with their existing backup infrastructure or purchase and manage state-of-art backup hardware and software. Without many options, some IT departments within the Premier League have responded to their shortage of expertise, hardware, and management skills by retaining the services of Backup Technology Limited (BTL) to help manage their data protection needs.

The Liverpool football club is a classic example of a team that had terrible problems managing its outdated tape storage systems which resulted in unreliable data restores and constant headaches managing its backups. Among its biggest challenges was completing backups in an acceptable backup window. Liverpool's rate of change for data and growing pool of data, especially when they factored in all of the video data that they needed to protect, made backup and recovery difficult to say the least. Exacerbating the situation, tapes were transported to an offsite location via a courier service. So if and when the Liverpool football club needed to restore data from those backup tapes, they first needed to place a call to that service.

It was into this environment that Backup Technology Limited (BTL) stepped in to lend the Liverpool football club a hand with its Information Recovery Management Platform - Asigra Televaulting.. With the use of Asigra's Televaulting SaaS platform, BTL was able to completely alleviate Liverpool's data protection pain points by providing:

Effective, affordable, and easy-to-manage solution for addressing the performance challenges caused by continuous growth in data
Agentless deployment and automatic self-healing with high backup and recovery success rates
Elimination of capital expenditures on storage infrastructure and wasted man hours

Following the success that Liverpool had with leveraging the leading backup technology solution available in the market that addresses all the strategic IT needs and gaining complete control of all backup operations, 7 of the other football clubs, including arch rival Everton, are now being protected by BTL's solution. Regardless of the exact issues each of these clubs had with backup and recovery, the cycle that lead to complete datacenter protection and adoption of the BTL solution was the same.

Initially a club would start with a few file servers to backup email and/or some key applications. Once comfortable and acclimated with BTL services, clubs begin to add a few more servers. Soon they would add their ticketing systems, fan databases, critical application servers, and then video analysis. All said and done, these clubs are now letting BTL protect every single packet of data that resides on their production systems and when new servers are procured, the clubs have the confidence to let BTL take immediate control.

Clubs within the Premier Football League require reliable and highly available backup and recovery solutions just like any other organization but face the same challenges - implementing a backup solution that is cost effective with reliable, consistent results. So in the same way that the Premier League broke from the Football League and rose to become the top ranked sporting event in the world, clubs within the League are breaking away from legacy backup methods. In so doing, they are looking past inter-team rivalries and taking a play out of the Liverpool club playbook by leveraging the benefits of BTL's Asigra Televaulting-based solution.

It's not Impossible to get all the Enterprise Backup & Recovery Features that You Need for Free

2008-12-02T11:00:00Z

Any storage architect or administrator that has ever dared to accept the challenge of engineering or re-designing their company's backup and recovery environment has undoubtedly discovered that he or she has had to sacrifice functionality or features based on the practical limits of their budget. Reasons for this vary from vendor to vendor, but mostly it comes down to how many backup and recovery software options are they willing to pay for? Most vendors offer reasonably good licensing for the core software, but once you step outside of that realm, some of the most basics features are not included. Some features I have seen omitted from what I now consider core functionality include:

Support for disk-based backups and VTL's (cost and quantities are very dependent on the specific vendor implementation)
Any type of Single Instance Storage, DeDuplication or Compression
Any Encryption technology, be they Hardware or Software
Advanced Server and Application level backup agents (Exchange, SQL, Sharepoint, Notes, Synthetic full support)
Bare Metal Recovery
Granular Recovery of a given applications set of data

This list is obviously not an all-encompassing one but it serves to demonstrate that these are features that are generally not typically included in the core software of the major backup vendors. However, it does demonstrate the need for backup architects to perform a significant amount of pre-planning and research to understand what they are omitting when they give up a feature for price if it's not included in the core software.

In my previous life, I have had experience in decisions like this, and they usually turn out less than positive for the end-user. About 90% of the time it was due to the vendor or the end-user really not understanding what they are losing when they give up a feature.

In that respect, wouldn't it be great if you could tear down those budgetary ramparts and say, "I want to cover my backup infrastructure end-to-end regardless."? Well, don't fret. There is a solution out there which oddly enough doesn't nickel and dime your budgets to death. Asigra TeleVaulting provides an all-encompassing backup and recovery solution, that provides all the layers of backup and recovery functionality needed to ensure that you can simply and easily achieve those backup SLA's without breaking the bank on licenses or manpower resources.

Asigra gives companies the flexibility to maintain this environment on your own, store the data to one of these emerging storage clouds, or even outsource it by using one of Asigra's many MSP's (Managed Service Providers) to assume the responsibility for the backups. Asigra Televualting includes many of the features for which other vendors charge for "free" as part of its core software offering. At the high level Televaulting includes:

Disk-Based system (No need to worry or deal with physical tape or multiple copies of those tapes, removes all those operational headaches)
Agentless operations (No need to manage nor maintain those pesky client/server agents)
Inclusion of space savings features (Deduplication for both DS-System storage as well as replicated backups with all data compressed after it is deduplicated)
Encryption functionality (AES 256 is included standard with the product)
Embedded compliance support (FCRP, SOX, HIPPA, PCI-DSS)
Full System Recoveries offered as part of the package (Standard system recovery procedures apply)
Granular recovery of databases as well as email stores (down to the message level, with the appropriate software)
CDP (TIVO for your infrastructure)

This is just a summary of the robust features that are included with Asigra's TeleVaulting product. So, the next time you are looking to re-architect or create a new backup and recovery environment, keep Asigra Televaulting in mind as it potentially offers some significant ROI and TCO for your corporation. Also, keep in mind that when you start to consider the trade-offs of sacrificing features that you need to maintain a robust backup environment because of their costly add-on features, there are unique solutions out there that can help side-step this concern.

Crafting a Proper Backup & Recovery Service Level Agreement

2008-11-17T11:00:00Z

Anyone who works as an end-user is continually confronted with crafting SLAs for various infrastructure components. Aggravating the situation, once SLAs are signed-off on, it is nearly impossible to make changes without completely rocking the boat so it is extremely important to get it right from day one.

First, you need to determine the scope of the backup and recovery SLA both in terms of what it will cover, and maybe more importantly, what will it not cover. It is crucial that as SLA objectives are crafted that they are clear and to the point to cover both yourself and your customers. To do this, you should have as many face-to-face meetings with your customers about the SLAs to ensure everyone understands the terms of the SLA and that they are in complete agreement about them. and total understanding.

Crafting the SLA objectives for your internal backup and recovery environment will need to cover any component of the infrastructure that may impact the successful backup or restoration of critical business data. These areas may include:

Backup Libraries (Disk-Based or Tape Based)
Offsite/Internal Media Storage Locations (How fast can you retrieve those tapes?)
Backup Servers (Masters or Media Servers)
Backup Reporting Servers
Client Backup Software (SAN Connected)
Backup Client Priorities (Business Critical Nature of the Application Server)
SAN Network (Fibre Channel, iSCSI)
Ethernet Network (Connections to Clients and Backup Servers)
Backup Operators
Data Center Environmentals

Once you have determined the appropriate infrastructure that should and should not be included in the SLA, then you need to determine the metrics that you will report back to your customer, to ensure you are delivering on all of the agreements defined in the SLA. The metrics are really determined by how much of the infrastructure and people either you or other IT teams are responsible for.

I encourage you, especially if you work in a large IT organization, to develop OLAs (Operational Level Agreements) with sister groups that you need to support as part of the overall backup and recovery environment (Windows & Unix Teams, Operations & Implementation Teams, Application & Database Teams, Facilities Team, and Outside Vendors were applicable).

Only when your infrastructure components, metrics, and OLAs are in place should you begin to put the whole SLA together. Depending on your customers' requirements, the SLA can vary greatly from nauseatingly detailed to simple and to the point.

One very effective way to accomplish this SLA engagement is by using software to track, monitor, and control every aspect of the backup and recovery infrastructure and process. One example of a product that does this is Asigra Televaulting. Taking advantage of Asigra's Televaulting backup and recovery software will give you and your organization the ability to define all of the various levels of the SLA and OLA processes and directly implement them in your backup and recovery environment. Asigra's software ensures that not only is the backup environment properly classified, but also that you the have the ability to customize the reporting and notification parameters around your SLA directly into the software. This greatly enhances your ability to meet and, in some cases, exceed your SLAs.

Anytime you can take the human factor out of a process including SLAs, and automate it via software, you have just placed yourself and your company in a much better place to be successful.

For those of you just establishing an SLA process for the first time, there are many sites to help you in this endeavor.. These include:

http://www.itil-officialsite.com/home/home.asp

http://www.service-level-agreement.net/

http://www.itsmsolutions.com/newsletters/DITYvol2iss14.htm

Overcoming the Agent Conundrum in VMware Environments

2008-10-24T10:00:00Z

When server and storage managers out there hear the "A-Word" (Agents) come up in a conversation with a software vendor, they typically cringe, and think to themselves, "Oh great, another set of agents that I have to not only deploy but that I have to manage and track." In the server world, some agents are unavoidable, like performance/security monitoring, virus and worm detection and prevention etc. However, most IT groups attempt to keep the number of agents that need to be installed on servers - physical or virtual - to a minimum, because for every new agent running on a server, that is yet another aspect of the server's configuration that must be validated and tested whenever any upgrades are performed.

I had an interesting conversation with a server manager at a Fortune 100 company last week and she brought up an interesting point about adding agents into her environment. Not only was she concerned with the items listed above, but also that the valuable server resources (CPU, Memory, Network, and Storage) were going to be hi-jacked away from the application when the server was put in service.

Now bring VMware into the picture when you are talking about having to manage those massive numbers of agents across 100's and possibly 1000's of virtual machines (VMs). The task can become a daunting one for any server manager to deal with even with the latest and greatest software management tools.

Backup and recovery services in most organizations requires an agent to be installed across all the servers in the environment that one needs to backup to tape, disk, or VTL. But just for a moment imagine if all those agents could be removed, not only from the VM's, but from the ESX servers themselves.

Asigra Televaulting's unique capacity licensing model means you no longer have to pre-pay for backup agents that you think you might need. For example, companies continually have to fight the "My standard server growth was X last year, so I need to budget for X more licenses this year" battle. Since Televaulting is licensed based on total capacity, you no longer need to forecast how many agents you will need which is only going to get more difficult in the future since with the advent of VMware you will no longer be able to see new VMs as they are created.

You also no longer need to install any backup and recovery software on VMs. Asigra Televaulting has always used an agentless backup and recovery process so still it does not require companies agents to install agents of physical machines to obtain full and complete backups. However it does need to take a different route to discover VMs as they are created on ESX servers. This is done as part of Asigra Televaulting 8.0's new integration with VMware's VirtualCenter. Using VirtualCenter, it can now discover and backup new VMs as IT administrators create them on specific ESX servers. This feature is also core to the Asigra Televaulting 8.0 software so companies do not need to pay extra to obtain it. Its capacity based licensing also stays in force so whether you have 100 or 1000 VMs, all that matters is how much capacity you are using to store the backup data from those VMs.

Data protection still remains at the top of many company's list of their biggest pain points and the need to install and maintain backup agents is one reason many companies remain hesitant to tackle this problem. Asigra's legacy of only offering an agentless backup solution coupled with its new integration with VMware give companies a near seamless way to implement data protection across their environment (whether physical or virtual) without the typical administrative pain associated with this type of change. In so doing, Asigra helps companies avoid the typical agent conundrum associated with new backup software deployments while solving a longstanding problem in their environment.

Real-time Data Protection without Host Agents: Is it Possible and Who Does it?

2008-10-17T10:00:00Z

The benefits that continuous data protection (CDP) technology provides as part of a company's overall data protection strategy are becoming more evident everyday. Point-in-time restores, faster recoveries and off-site replication of data for disaster recoveries are just some of the benefits that companies using CDP are already experiencing. However one of the challenges that companies may encounter as they look to deploy CDP that may hinder or even prevent its adoption is the need to deploy host agents on servers.

Deployment of host agents is almost always a prerequisite when protecting Windows or Linux servers as they are needed to handle one of CDP's unique requirements: copying every host write I/O. Without this host agent, the only other way to copy write I/Os is to attach host servers to either a storage area network (SAN) or a storage system that supports CDP. That usually limits CDP to high-end application servers and even then companies may still need to deploy agents to obtain the functionality they seek.

Yet today's requirements for corporate data protection for all application servers are increasing so companies now expect near real-time protection and recovery regardless of what type of server it is. In fact, it can be argued that the need for near real-time protection and recovery on Linux and Windows servers is as great as for high-end application servers since there are so many more Linux and Windows servers in corporate environments and fewer administrators to manage them. Since administrators may not know specific application recovery point objectives (RPOs) or when these objectives change on these servers, they can use CDP as a means to dynamically meet the changing requirements of these servers. Further, CDP can reduce the amount of time and administrator intervention needed to do such recoveries.

The trade-off is that companies usually need to deploy CDP host agents on their Linux and Windows servers to achieve these results. However there is one notable exception to this rule: Asigra Televaulting. While Televaulting's primary claim to fame is that it can perform agentless backups, what is not so well known is that it can also agentlessly protect servers using its built-in CDP feature. The exact way that Asigra does this is part of its intellectual property and it does not publicly disclose those details but at a high level it works as follows:

The DS-Client (Asigra Televaulting's local backup server) registers itself with the target server it is going to backup so it is notified of changes to files or emails that it hosts.
Whenever the DS-Client receives a notification from the target server that a change to a file or email has occurred, it queues the specific file or email on the target server for backup.
The target server then transmits that file or email to the DS-Client. Since the DS-Client deduplicates the file or email, it only stores the changed blocks of data in that email or file.
Finally, the DS-Client compresses and decrypts the file or email before sending it to the DS-Server (Asigra Televaulting's central backup server) for off-site data protection and recovery.

Using CDP to deliver real-time data protection at the email or file level with near-real time recoveries is something that many Linux and Windows administrators may crave but view as out of reach when they contemplate the need to install agents on hundreds of servers (physical or virtual). Asigra Telelvaulting overcomes this objection. Since CDP is core to Televaulting's software offering, companies receive it when they select Televaulting while eliminating the need for the deployment of host agents on servers. Though companies still need to verify that email or file level CDP is sufficient for their environment, Asigra Televaulting provides companies a transparent way to introduce CDP into their Linux and Windows environment while helping them achieve the higher levels of protection and recovery for these application servers that many companies are seeking.

Data Protection for VMware VMs is the new Invisible Problem; "They Reproduce Like Rabbits"

2008-09-29T10:00:00Z

If you do not yet think that virtualization is a real phenomenon, one only needs to attend the VMworld conference that recently took place at The Venetian in Las Vegas, NV. While VMware did not disclose the exact number of people in attendance, it was rumored that over 14,000 users and 2,000 vendors were in attendance. Gauging by the throngs of people flooding the sessions and exhibit floor, those estimates seemed reasonable. Yet what brought many of these attendees to VMworld was not only to learn more about VMware but how to better manage the new environment that virtualization creates.

A recurring theme in terms of what I hear from users is how VMware adds new complexities to their day-to-day management tasks. For instance, even before server virtualization came in vogue, companies were already complaining that their physical servers reproduced like rabbits. Server virtualization just makes server growth that much easier to occur since now companies don't even need to purchase a new physical machine anymore - it now is little more than a copy-and-paste like exercise to create a new virtual machine (VM) once server virtualization is in place.

Adding to every company's concerns, server virtualization is still in its infancy. In fact, at this fall's Infiniband Trade Association (IBTA) Tech Forum that preceded VMworld, Gartner's John Enck estimated that companies have only virtualized about 7% of their servers. By 2012, he forecasts that the number of servers virtualized in corporate environment will grow to 60% or more.

It is also natural to conclude that the management problems and levels of complexity associated with managing virtual servers are minor right compared to what they will be once server virtualization fully takes hold in corporations. Nowhere does this hold truer than when it comes to protecting VMs on physical machines. Aside from the performance overhead that backups can incur on the physical server when they run, a thornier problem many first face is identifying what VMs exist on a physical server as well as when new VMs are created. Since server administrators are not always the same individuals as the backup administrators, unless companies have some mechanism to track when a new VM is created and schedule it for backup, the distinct possibility exists that it will never be backed up.

Recently at least one data protection product, Asigra Televaulting, has come out with a mechanism to deal with this escalating complexity in data backup. While Asigra Televaulting has always provided companies with the means to do agentless backups of physical machines, to detect and backup multiple VMs on a single physical machine adds a new level of complexity even when an agentless mechanism is utilized.

To discover VMs in this new virtual infrastructure, Asigra Televaulting communicates with VMware's management server, VirtualCenter, to obtain a list of all VMs on each physical machine. In so doing, it not only can identify what VMs are already on each physical machine, but alert backup administrators when new VMs are created so these VMs too can be added to the backup schedule. Alternatively, administrators can also create a backup schedule that automaticallys adds any new VMs if and when they are created so they are not missed.

Corporate adoption of server virtualization is already a real phenomenon and accelerating with each passing day. However companies need to recognize that as they bring server virtualization in-house they also need to account for how they will identify existing and new VMs as they appear on physical machines and then protect them. The enhanced integration that Asigra Televaulting 8.0 now has with VMware VirtualCenter not only makes the detection of existing and new VMs possible, it helps to remove some of the management complexity that the introduction of virtualization creates.

Disk-based Backup is more than just Backup Relief

2008-09-17T18:00:00Z

Companies can experience an overwhelming sense of relief when they finally resolve their ongoing backup problems by switching from tape to disk as their primary backup target. But what companies may fail to fully contemplate is the new possibilities - and challenges - that storing data on disk opens up to them. On the upside disk makes data recoveries and off-site replication of the data much easier to accomplish. Conversely, it can present companies with new challenges to manage the data on disk as it ages lest the escalating costs of disk capacity and cooling and powering the storage system start to offset some of the benefits that disk-based backup provides.

In previous blog entries I have examined how Asigra's new Backup Lifecycle Manager (BLM) Archiver feature in its Televaulting 8.0 release helps control the placement of backup data. The BLM Archiver optimally places aging data on the appropriate tier of disk without sacrificing the ability of companies to recover their data. It keeps recently changed or new data on higher performing disk on its DS-Systems while moving infrequently or never accessed data to more energy efficient storage systems used by its BLM Archiver. This keeps costs low while still keeping the backup data on disk so companies still have the flexibility to manipulate and search backup data over time in the manner that they so choose.

The most pressing need for these capabilities is to satisfy emerging corporate needs for eDiscoveries and the subsequent legal holds that are on the rise especially in the US due to changes introduced as part of the updated Federal Rules of Civil Procedures (FRCP) in December 2006. As Asigra Televaulting initially backs data up using its agentless architecture, it also indexes data that is backed up and records such attributes as file name, file size, file create and modify times and directory name. When backing up emails, it tracks email information such as the email subject and recipients (To, From, Sent, CC, BCC, Sent and Received). This information remains with the backup data as it is archived by the BLM Archiver so companies can search the BLM Archiver and retrieve the archived backup data based on these criteria.

Companies may also find that they need to retain some or all of the data that matches the search criteria of the e-discovery and place a legal hold on it. Using the BLM Archiver, companies can copy the specific data that they need to retain to satisfy the legal hold a specific tier of disk (such as WORM based disk) or even copy just that subset of data off to optical or tape if the requirements dictate.

In the event that a more granular search of backup data is requested (such as searching the content of files or emails) at a specific past point-in-time, using the BLM Archiver companies can first identify all data on that date and then create a restorable image from it. Once the backup data is restored, companies can then do this search across the files and email for that date.

BLM Archiver's ability to copy data to multiple types of media also comes into play in other situations. The BLM Archiver can recognize and store data to any type of storage device to include file servers that use CIFS and NFS interfaces. This allows companies to store their archived backup data stores to emerging storage platforms that are based on new cloud computing and grid storage architectures.

The problem today is not backing up data to disk, it is managing the backup data once it is stored there. Asigra's BLM Archiver not only gives companies a cost effective mechanism to manage backup data, it helps companies to search and hold data in their backup stores while still giving them the flexibility to take advantage of emerging and new storage technologies. In so doing, companies can expeditiously respond to unexpected legal ediscovery and hold requests, not put undue burdens on their IT staff and still keep their disk storage costs under control.

Data Destruction is also part of the Lifecycle of Backup Data Management

2008-09-10T18:00:00Z

Asigra makes no bones about it: it unabashedly advocates that companies keep all of their backup data on disk under the management of its Televaulting software. The reasons Asigra provides for keeping backup data on disk are plentiful as well: Faster backup and recovery times; elimination of tape management tasks; deduplication technologies that minimize data storage requirements for disk; and, data that is easy to copy and replicate locally and remotely. Yet if there is anything companies know about backup, it is that managing backup data and its recovery over the long term, whether it is on disk or tape, is where the complexity can start to surface.

In a recent blog entry I examined why Asigra introduced the Backup Lifecycle Manager (BLM) Archiver feature into Televaulting in order to optimize the placement of aging backup data on different tiers of disk. But the BLM Archiver does more just simply move data out of the Asigra DS-System. The BLM Archiver also carves the disk that it manages into "Staging" and "Consolidated (Disk Based)" buffer areas as shown below:

The Staging and Consolidated buffer areas within the BLM Archiver are automatically managed by the BLM Archiver so companies need to do little or nothing with these two areas in terms of day-to-day management once the BLM Archiver is configured and operational. However these two buffer areas within the BLM Archiver help to save money as well as manage the data once it is archived.

The practical application of these two staging areas initially comes into play as data is moved from the DS-System to the BLM Archiver. As data is moved into the BLM Archiver, the Staging buffer area accumulates incoming files from the DS-System over a period of time. Since the DS-System only sends obsolete data to the BLM Archiver, the influx of obsolete files to the BLM Archiver typically does not happen in bulk. 100 files may come in one day, 5 files the next day, 50 the next and so on.

In order for the BLM Archiver to ease the management of these smaller incoming files over time, it will create one large consolidated file that contains all of these smaller files. The creation of this large file only occurs when the Staging buffer fills up. The default size of this buffer is set by the user and it is usually set to at least 2 GB in size but can be larger. Once the Staging buffer fills up, these small files are consolidated and packed into a large file. At this point, the large file is moved to the Consolidated buffer. Once moved, the BLM Archiver guarantees these files (small and large) will remain stationary and not be touched again unless a restore is needed.

A situation where companies may need to restore files from the BLM Archiver is if they need to recreate what a specific server looked like at a specific past point in time to satisfy a legal discovery or hold request. In this situation, a "recoverable image" may be needed. To create this "recoverable image" of the server in question, the BLM Archiver gathers needed files from both the Staging and Consolidated buffer areas to create and present a "recoverable image" that could be used for the search. In the event companies need to keep this specific snapshot of data on another type of media (disk, optical or tape) to satisfy a legal hold, they can also copy this "recoverable image" to the media at this time.

From restorability and searching perspectives users will see little difference regardless of which buffer the files are stored within the BLM Archiver. Any search will automatically search both the Staging and Consolidated buffer areas since restorable images may consist of files from either buffer. The only difference a user may see is in the Televaulting GUI which displays the status of the file ("Staging" or "Consolidated").

The more important difference between the Staging and Consolidated buffer areas has to do with managing the destruction of this obsolete data on disk. Companies may only want to keep this archived data three to seven years. Using the BLM Archiver, companies can set policies on each large file as it is created so that the BLM Archiver destroys the file once it expires.

Most users will essentially find the introduction of the BLM Archiver into an existing Televaulting environment a "Set it and Forget it" exercise. It gives companies the assurance they can search and recover archived data should they ever need it to satisfy some unforeseen future legal eDiscovery or hold request. More importantly, companies can also use the BLM Archiver to protect themselves from over compliance by deleting data that they are no longer bound by internal or external laws to keep. Yet a component of the BLM Archiver story that needs further examination is the options it provides for searching the data already in the archive. I'll cover that material in a forthcoming blog entry.

Asigra Embeds New Encryption Key Management Options into Televaulting 8.0

2008-08-27T18:00:00Z

When trying to decide on an encryption strategy there are numerous areas of concern, but encryption key management is one area that should heavily influence whatever your strategy ends up being. Some important areas of consideration when making the selection include evaluating how the solution handles:

Static and dynamic key generation
Key management technique used to encrypt and decrypt the data
Administrative overhead associated with the key management

Since encryption is only as strong as your key management, it is important to take into account each of these areas in order to determine how to best implement encryption as well as its corresponding key management solution.

In Asigra's recent release of Televaulting 8.0 data security remains at the forefront with their use of the AES encryption algorithm to encrypt data while in transmission across the network; or at rest in its DS-System or BLM Archiver. Televaulting's approach to encryption key management provides several options in how to best approach encryption key management. Televaulting 8.0 gives users and service providers several key ways to protect data from unauthorized exposure:

Starting with Televaulting version 8.0 copies of encryption keys can be kept on the DS-System. The DS-System keeps the keys of the DS-Client in escrow, but in order for this feature to be functional both the DS-Client and the DS-System has to be enabled to "forward" and "accept" the encryption keys. Having the ability to store keys on the DS-System lowers the chance that an encryption key could be lost and thus making data unreadable.
The DS-System can be set to "encryption key escrow mandatory". Using this setting, the DS-System will only accept DS-Client connections after they forward their encryption keys. This provides service providers the assurance that DS-Client keys are kept in escrow lowering the possibility of key loss.
Keeping encryption keys in escrow could allow a liability risk for service providers. In Televaulting 8.0, Asigra allows service providers to accept encryption keys into escrow only if the DS-System is enabled to do so. Without enabling the DS-System to escrow encryption keys, the DS-Client will not be able to forward the keys to escrow.
DS-Client encryption is always mandatory. Providing mandatory encryption ensures data at rest is encrypted, as well as when it is in transmission to the DS-System. This type of encryption arrangement ensures legal compliance if an unauthorized exposure of data should occur.
Encryption keys stored on the DS-System can be forwarded to the BLM Archiver. This type of transfer for archival purposes is paramount to data recovery in the event of a disaster.

When setting up Asigra's Televaulting 8.0, upon installation the DS-Client is required to create an encryption key thus generating a static encryption key. Proper controls over the DS-Client encryption key should be used so as to lessen the chances of unauthorized key exposure. Third party encryption appliances and key management systems are not directly supported, but if one is used in the network background it is completely transparent to the DS-System, DS-Client, and BLM Archiver.

When deciding on a backup solution, especially when data is at rest or in transmission is concerned, it is important to understand what role encryption and key management plays in the chosen technology. Understanding whether static or dynamic key generation is needed or preferred depending on the level of security needed for your environment, how best to manage encryption keys, the associated administrative overhead, and liability of lost or exposed encryption keys should all be considered. By providing flexible choices in how to manage both encryption and encryption keys while allowing for data and liability protection, Asigra Televaulting 8.0 has struck a balance to meet both data center and service provider needs.