That means that the amount of enterprise data that has to be protected -- backed up or otherwise replicated -- continues to grow and grow as well.
Indeed, many enterprise backup environments put in only a handful of years ago are already feeling the pain and are thus likely candidates for a major fresh look.
It's in this fast-moving market that EMC's BRS (backup and recovery systems) division continues to innovate and thrive.
And, as part of EMC World, they've done their part in announcing important new capabilities: some modest, some quite large :)
A Market In Change
I'm not going to say "tape is dead", but it's a certainly not a growth market, either …
EMC plays in this market two ways: hardware and software. Both are (of course) largely agnostic to the other, but they tend to work very well when used together.
From a hardware perspective, our lead product is Data Domain. They've just about run away with the PBBA (purpose-built backup appliance) market -- the modern successor to the tape library market.
More importantly, we've been able to help many of our customers achieve an important strategic goal for them -- going tapeless.
We've now got over 1,000 customers who can scream loudly and proudly "I'M TAPE FREE!!!" for operational recovery, including some surprisingly big names.
The New Big Box -- the DD 960
Years ago when we were wrestling NetApp for the Data Domain acquisition, one of the key points we were forced to consider was Data Domain's underlying product architecture.
They had done a great job architecting deduplication around Intel's multi-core roadmap, and -- well -- our own product guys hadn't exactly done that. Ouch.
That architectural investment meant that every time Intel came out with a faster chip and more cores, the Data Domain product automatically got bigger, faster and cheaper with comparatively little engineering effort.
We at EMC didn't have that sort of architectural advantage. The rest is history, as they say ...
Years later, you can now see the effect in the Data Domain product offering. Last year's "big box" was the DD 860 and 890 -- no slouches at the time.
This year, it's the DD 990 -- arguably the biggest, fastest and hence most efficient backup target in the business. Not by a small margin -- but by a considerable margin. I think it's a fun game to simply predict the envelope of the next DD 'big box' by simply extrapolating past efforts based on the Intel's expected roadmap.
Note that the team makes a point of emphasizing the inline nature of the data deduplication. Although there are approaches out there that dedupe asynchronously, they can be a handful to manage, especially in demanding enterprise settings. Most serious consumers I've found now prefer the inline approach.
More Boost In More Places
The DD Boost technology works by off-loading some of the compression activities to either the application server, the backup server, or both. A neat capability that can dramatically improve performance at not much extra cost.
With this announcement, there's Oracle RMAN support, but -- more interesting -- DD Boost support for Greenplum databases used in big data analytics.
You might have been hoping that you could get away without having to backup petabytes of scale-out databases, but -- of course -- no such luck. Big data needs to be backed up, just like the other kinds.
Bring some fast kit, otherwise you'll be there for a while :)
Avamar 6.1 Enhancements
I've always been a fan of Avamar -- client-side dedupe data protection with an architecture that's achingly elegant. If there's any sort of a network involved between your backup source and target, you'll be interested in what it does.
With this release, it's more of a good thing.
Avamar updates support for the all-important Oracle and SQLserver use cases, but also adds very important support for both SAP and Sybase.
In particular, there's a new multi-streaming architecture that can dramatically improve performance of both backups and recoveries in virtualized environments.
As part of the announcement, there's extended retention support for both Data Domain and Avamar, including a Lock Compliance feature for Data Domain if you retain regulated data. In particular, the Data Domain Archiver product announced last year is now known as the Data Domain Extended Retention Edition (a feature vs. a distinct product)-- now with extended capacities, and new support for the IBM i series.
More details are in the extended announcement materials if you're interested.
EMC -- The Data Protection Leader?
It's an important business, we do well at it, and we continue to invest -- as past success is no guarantee of the future.
And that enviable market position gets granted to us before we get to data replication (SRDF, RecoverPoint, et. al.) or even the object-oriented protection schemes found in Atmos and Centera.
Or the data protection management products like Data Protection Advisor, Replication Manager and the new AppSync.
Historically, many of our competitors have offered the view that our portfolio is way too complex -- we ought to have only one product and only one way of doing things.
I suppose that's what you'd say if you only have one product to offer.
That being said, I think we'll keep doing what we're doing -- although there's always room to improve, it's working out pretty well for us and our customers.