The Right Architecture Is Priceless, Part V

Stephen Manley

Stephen Manley

CTO, Data Protection and Availability Division
Over the past 15 years at both EMC and NetApp, I have traveled the world, helping solve backup and recovery challenges - one customer at a time (clearly, I need to optimize my travel arrangements!). My professional mission is to transform data protection so that it accelerates customers’ businesses. I have a passion for helping engineers pursue technical career path(without becoming managers), telling stories about life on the road and NDMP (yes, that’s NDMP).


The John Hancock Tower (a.k.a. The Hancock) in Boston is synonymous with architecture disconnected from reality. A 60-story glass skyscraper in Boston, the tower is a tourist attraction and an exemplar of modernist architecture. But it also has a long, embarrassing history. At one point, it was dubbed the “Plywood Palace.” When 500 pound panes of glass fell off the building, engineers temporarily replaced much of the glass with plywood. Then, in high winds, the tenants of the upper floors suffered from motion sickness; to solve the problem, engineers installed a tuned mass damper. Finally, when it was discovered that the building could collapse under very high winds, engineers added thousands of pounds of steel bracings.

A monolithic architecture can look elegant on paper, but real life engineering depends on flexibility and the ability to evolve quickly. The right architecture needs to translate into reality.

Traditional Backup vs. Protection Storage Architecture

A perceptive reader asked: “How is your Protection Storage Architecture different than a traditional backup architecture where backup software (management) triggers backup clients (data sources) that send data to tape (protection storage)? And why should I care?”

Traditional backup architectures are closed and monolithic. The Protection Storage Architecture is open and modular.

  • Data Access: Traditional backups store data in a closed, proprietary format; you cannot access your data without the backup app that created it. The Protection Storage Architecture prefers to store backup copies in their open, original format – i.e., backups of database files as database files, VMs as VMs, and files and files – instead of locking the backups into proprietary tar images.
  • Metadata Access: Traditional backups limit backup visibility by closing off access to the protection metadata. Non-backup administrators, so they cannot easily understand what the backup application is doing. The Protection Storage Architecture opens backup visibility and control to the data source owners, via their preferred interface (e.g., Oracle RMAN, VMware vSphere).
  • Alternate Protection Methods: Traditional backup architectures refuse to manage or monitor backup copies created outside their monolith (backup application driving backup client to write data to backup storage). They also struggle to support new data source layer performance optimizations (e.g. snapshots, replicas, changed block tracking), delivering a sub-par bolt-on module years later. The Protection Storage Architecture’s modularity enables it to treat all protection copies as 1st class citizens. This includes all types of protection (e.g. snapshots, replicas, backups, archives), created by all users (e.g., DBA, VM admin, or any third-party backup software), and all optimization modules.

The difference in architectural approach is stark: closed and monolithic vs. open and modular.

What’s the Value?

The value of the Protection Storage Architecture boils down to two things: performance and visibility.

  1. Performance: A traditional backup client cannot scale backup and recovery performance with data growth. On backup, it reads through all the data to find what to protect and write into its proprietary format – searching for needles in haystacks. On restore, it must translate all the data from the proprietary format and write it to primary storage. The data can only be accessed once the customer finds primary storage capacity and runs a, potentially, multi-TB recovery.A modular architecture leverages the intelligence in the data sources (e.g. hypervisor, application, primary storage) to optimize protection performance. Instead of searching for needles, the data sources can track exactly what new data to protect because they’re writing the data. One can reduce backup and recovery times from hours to seconds… if your architecture is modular enough to leverage the intelligence in the data sources.The Protection Storage Architecture scales recovery because it leaves data in open formats. Since it stores data on disk in its original format, customers can instantly access their data. In the event of a disaster, an application can be up and running in minutes, instead of days. Or, if multiple users lose files, they can recover their own data, instead of hitting a bottleneck waiting for the backup team and tools to help them.The open, modular architecture optimizes and scales both backup and recovery performance.
  2. Visibility: With traditional backup architectures, everything must run through the backup team bottleneck. These architectures drive a “my way or the highway” approach from the backup team – my backup app, my schedule, my clients, my data format, my absolute control. It’s no surprise that other groups are choosing the highway and rolling their own solutions.The Protection Storage Architecture wants to increase everybody’s data protection visibility. Regardless of how a protection copy is made, or by whom, the Protection Storage Architecture’s Data Management Services will discover, report on, and catalog it. Furthermore, the Protection Storage Architecture wants to ensure that each key user has a native user-interface into all versions of their protected data – be it their application (e.g., Oracle, SAP), hypervisor (e.g., VMware), storage array (e.g., ViPR), or user protocol (e.g., NFS/CIFS).

How Do I Adopt the Protection Storage Architecture?

One of the virtues of a modular architecture is that you don’t need a “rip and replace” to move to the architecture. The modularity of the Protection Storage Architecture enables companies to set long-term goals, while deriving value today and each step of the way. How you move forward depends on your environment, your business and your objectives. Some choose to adopt whole layers (e.g.,protection storage or data management services) across their legacy environment; others opt to focus on end-to-end use cases (e.g. database backups, VM backups), and still others decide to follow a hybrid approach.

Your organization, like ours, is moving to a services-oriented backup and IT approach. As it does, it’s critical that your foundation is open and modular.

Backup is broken because the traditional architectures are closed and monolithic. Companies will not succeed in deploying rich services on top of a legacy backup architecture. With the Protection Storage Architecture, the engineering team can evolve the environment, quickly solve pressing business challenges, and unify data protection services.

Amazingly, in 1977, just after the plywood debacle, the American Institute for Architects bestowed the Hancock Tower with a National Honor Award. Sometimes, architects forget that the right architecture is not about a beautiful monolith on paper – it’s about putting the engineering team in the best position to solve real-world problems.

The Right Architecture Is Priceless, Part IV

Stephen Manley

Stephen Manley

CTO, Data Protection and Availability Division
Over the past 15 years at both EMC and NetApp, I have traveled the world, helping solve backup and recovery challenges - one customer at a time (clearly, I need to optimize my travel arrangements!). My professional mission is to transform data protection so that it accelerates customers’ businesses. I have a passion for helping engineers pursue technical career path(without becoming managers), telling stories about life on the road and NDMP (yes, that’s NDMP).

120893069Architecture is priceless, and history wants to make architects pay dearly. According to legend, after the completion of St. Basil’s Cathedral in Moscow, Tsar Ivan the Terrible gouged out the architect’s eyes; he did not want him to design a superior structure. Similarly, when the Taj Mahal was completed, Shah Jahan supposedly cut off the architect’s hands. While historical evidence indicates these are myths, architects should consider in the long term, what value can they deliver to keep themselves safe?

In this series, I introduced the Protection Storage Architecture , explored Protection Storage and Data Source Integration. Now to Data Management Services, where the fun begins.

Data Management Services – Why Do They Matter?

What value should a protection team deliver? First, application teams need to access the right version of data, when they need it. That means high-performance and user-visible data protection. Second, management needs analytics and reporting for regulatory compliance, service-level enforcement, and cost management. Third, users, auditors and lawyers need to be able to access the right historical information, quickly and accurately.

Without data management services, companies fail audits and lose data, productivity and business. Only the backup team has the technology and organizational scope to deliver the protection services.

Data Management Services – Why Have They Become a Challenge?

A decade ago, the backup application exercised absolute control over every aspect of the end-to-end protection workflow. Therefore, it was simple for it to deliver reporting, cataloging and management of protection storage (a.k.a.  tape). Things changed when traditional backups could not scale with data growth and virtualization. To meet their business needs, the application, hypervisor, and storage teams began to deploy their own solutions, using tools integrated into their data sources – the accidental architecture. The backup application (and team) lost control and visibility over the environment.

Despite the increasing variety of protection mechanisms, the business still needs consolidated data management services. The backup team must gain visibility into all the protection copies – even those created outside the backup application.

Data Management Services – The Baseline

The protection team must ensure:

  • Recovery and Access. Information must be protected, regardless of: where it resides (data center, remote office, mobile device, cloud), the protection mechanism (snapshot, data-source mover, backup or archive client), and the location of protection copy (backup appliance, cloud, or tape). The protection data must be accessible, regardless of use case (disaster recovery, operational recovery, archive access), mechanism (instant access, granular object recovery, or search-driven access) or who drives the recovery (protection team, application admin, end-user). Even though the backup team may neither create nor recover the protection data, the business still holds it responsible for rapid and reliable recovery and access. (And, yes, backup teams, now is the time to curl up into the fetal position.)
  • Policy Management.  The protection team must create, manage, and report on protection policies for both regulatory and service-level compliance. In addition to cataloging and reporting on all the copies, the protection team must supplement the data source administrator’s policies. For example, imagine that a DBA runs her own database dump and plans to retain it for 30 days. If corporate policy is 90 day retention, the protection team must find a way to deliver that retention. Ideally the supplement would be efficient (clone the DBA’s copy) vs. inefficient (create your own primary backup copy). Regardless, the protection team is the “one throat to choke” for policy compliance.
  • Protection Storage Management. Protection storage management involves far more than managing purpose-built backup appliances. This includes: primary storage (e.g. snapshots and replicas), backup appliances, WAN network bandwidth (to replicate between sites), tape and cloud. The protection team needs to ensure that the entire infrastructure is available for recovery and backup. Furthermore, they need to plan for the expansion of the environment – projecting future costs and purchases.

Data Management Services – Bonus Levels

With a consolidated view of the organization’s information, protection teams can deliver value beyond the baseline and accelerate the business. With data and metadata analytics (like the NSA of the enterprise) the protection team can expand their business impact.

First, organizations crave visibility and control over their infrastructure spending and growth. The protection team has can track data growth rates by division, application, location and data type. Not only does this help them project what they will need (protection storage, bandwidth, compute), but it can also provide insight into the primary storage and application environment.

Second, the protection team can derive insight from the data. Most groups either view only their silo of information, so they lack the full visibility across all the data. The protection team has a catalog of every piece of data in the environment. With that pool of data, some groups are already searching for security/compliance leaks, optimizing test & development and hardening their systems.

After delivering the baseline, the protection team is well positioned to deliver advanced services.

Protection Storage Architecture – The Value

The protection storage architecture positions the backup team to evolve into a high-value service provider. You need all three components of the architecture to succeed. You can’t provide the data management services without consolidating the protection. You can’t consolidate the protection without integrating with the data sources. And you can’t integrate with the data sources without the right protection storage. Hence, you need an architecture that loosely couples all three architectural components.

With the right architecture, the protection team can deliver tremendous and keep their eyes and hands.

Get Ready to Backup to the Future on 7.10.13

Mark Doncov

Mark Doncov

I’ve spent most of the seven years I’ve been at EMC on backup. Currently, I work on category and solution marketing initiatives in EMC’s Backup Recovery Systems division. In short, this means I focus on the “why” for EMC Backup, not the “what”; I leave the bits and bytes to the product teams. Over the years, I’ve seen big changes in the backup world. I will be looking at these – and the even bigger ones that lie ahead - here on The Backup Window and other social channels.

Does your backup make you feel like this, spinning uncontrollably?

It’s easy to see why.  Data growth is unrelenting.  And today’s backup can’t cope with the scale, complexity or costs that result.

On July 10, 2013, we’ll reveal a new generation of backup and archive breakthroughs designed to get you out of free fall – and in control.

EMC President and COO David Goulden and Backup Recovery Systems division executives Guy Churchward and Stephen Manley will be the guest speakers.

There will be live Q&A.

Oh yeah – there will also be the 5,280 foot drop of death you got a glimpse of.

Watch the future of backup unfold for yourself on July 10th – just click here to learn more or here to save the date to your calendar.

And make sure to read all about it next week on The Backup Window.

See you next Wednesday!

EMC Backup Recovery Systems – Video Blog from Cisco Live

Chandra Jacobs
I love creative and challenging projects in the emerging technology product space. I have a background in tech, innovation, and product development, especially as applied to web and mobile apps in the entrepreneurship arena, but have recently moved into marketing. In my role as a product marketer, I have gravitated toward digital marketing as well as analytics/data mining. It fits well with my techie geek bent as well as my cloud angle on The Backup Window. (Be sure to catch my posts on Innovation Station too!) Outside of work at EMC, I enjoy exploring Boston’s culinary and jazz scene (often in combination), and travel as much as I can (35 countries and counting).

Three Signs Your Company May Be Losing Its Mojo

Heidi Biggar

Heidi Biggar

Marketing and IT Consultant, Data Protection and Availability Division
I’m often asked how a political science major at Tufts wound up in the IT world, covering backup, storage, virtualization and cloud of all things. Truth is, it’s really a love for learning, a need to understand the “bigger picture” and a desire to share that view with others that’s steered my path over the past 20 years, from campaign manager to editor, analyst and marketer. After hours, you’ll find me hanging with family, running 10ks through Peachtree City’s 90 miles of cart paths, watching football or reading. I’m a New England transplant enjoying life in the South. In my previous life, I also blogged for ComputerWorld, Enterprise Strategy Group and Hitachi Data Systems, but The Backup Window is my baby. It's been great watching it evolve.


“You used to be much more… ‘muchier.’
You’ve lost your muchness.”
– The Mad Hatter to Alice in the film Alice in Wonderland.

Really, what could be worse than losing your muchness? Your mojo?

In the IT world, it can be the death knell of your business, signifying an impending downward spiral or deceleration.

There’s healthy “stick-to-it-iveness” and, well, then there’s, as Rich Castagna, editorial director of Storage magazine puts it in his May editorial, doing something [the same old way] “because that’s the way we do it.”

One’s okay… the other not so much. One keeps your business moving forward and in new directions… the other slows it down, and limits opportunities.

The good news is research data continues to show that more of you are rejecting the notion that doing nothing or the same-old-thing is a good thing … for your organization or for your own professional well-being, for that matter. You see the bigger picture (pun intended!), and you are transforming your backup environments and your thinking.

#1: Your company still views backup as a cost center

As a barometer (of the change that’s occurring in the market today), we polled the live Backup Game Day audience at EMC World last month, to see what’s really going on in customer environments. It’s a good reality check for us, and a good opportunity for users to see where their peers are at. Plus, it’s just a fun way to engage with you.

Anyway, we asked a bunch of questions, but one of the ones that stood out was, “How does your company view backup: as a cost center, a tactical necessity (to protect against “what if” scenarios), a game-changer (i.e., a strategic investment in their company’s future or evolving (i.e., somewhere in-between a tactical necessity and game-changer)?”

 Not (too) surprisingly (after all, change is tough, particularly in backup circles), the majority of Game Day respondents said backup was still viewed as a technical necessity. However, 26% said it was a game changer, and only 1% said it was nothing more than a cost center. So, I’m curious, where does your organization fall within this spectrum? Comment below, and let us and your fellow TBW readers know. And if your organization is a straggler, what’s holding you back?

#2: You’ve never heard of PBBA

Short for Purpose-Built Backup Appliance, PBBA adoption is another very good barometer of change, and according to the the just-released IDC Quarterly Worldwide Purpose-Built Backup Appliance Tracker, the market is still red hot! Revenue, capacity and shipments are all up significantly year-over-year, and at EMC still is driving the market, with more than 4x the market share our nearest competitor.

So, what about you? Got PBBA? If you don’t, the following IDC paper is a great level-setting resource:  Backup and Recovery Changes Drive IT Infrastructure and Business Transformation. Also, be sure to read Stephen Manley’s blog series: The Right Architecture Is Priceless. In that series, Stephen talks about the role Protection Storage (a.k.a., PBBA) plays in a transformed backup environment.

#3: Your CIO is hyper-focused on saving a buck

Times they are a changing, and so must CIOs.

If you’re a regular follower of The Backup Window and my posts, then this may sound familiar. Click on this link, and it’ll bring you to a post  I wrote back in the spring.

The bottom line is that for transformation to be successful, it’s got to start at the top – and extend down and across the organization. This means CIOs need to be 100% onboard with the need for and reality of transformation, and it means re-setting priorities so they’re in line with a services-oriented mindset.

And this is happening…

Of the more than 1,500 CIOs and other IT leaders, IDG surveyed as part of its recent CIO Research study, 49% ranking improving IT productivity as their number-one goal for 2013, followed by better, faster, decision-making; improving service levels; protecting corporate data and increasing agility. As for reducing costs – again, the traditional front-runner – it ranked a distant eight in the survey.


So, what about your organization? Has it lost its muchness? And what about you?