Executing on fundamentals: Simplicity and Efficiency

So, here we are at day 2 of EMC world.  If day 1 was all about “big picture” innovations in entirely new areas (VPLEX, IaaS efforts), today it’s about continued execution on the fundamentals for EMC.

To set the stage here a bit:

• At the high-end, VMAX and its next-generation capabilities and architecture is killing it.   Customers are voting with their pocketbooks, and the growth we saw in Q1 was fantastic – 28% year over year, and quarter over quarter growth that means we’re doing something right.
• VPLEX represents a new “category creator” – transactional storage models that can be federated locally and across distance.
• VCE and other transformative initiatives (see my post on IaaS here) is accelerating like you wouldn’t believe (more on this later).   Cisco, VMware, and EMC literally can’t hire fast enough to staff against demand.   I was in QBRs this past week and in some meetings with Capellas, and ‘Zilla nails it in his post Contemplating Capellas.  Again, customers and partners are voting with their wallets.
• Whole new fronts are opening up.   Sometimes people are confused by this (they want to “single category” vendors) – but remember that you need to fight and innovate on several fronts simultaneously otherwise you get  out flanked.   

So, what’s an example of ONE new front? 

OK, before we get to the details of today, I want to highlight how much it’s important to play on multiple fronts (while executing on fundamentas).   So… here’s an example.  Customers are telling us that no matter what they do, for some (not all) database applications, they simply can’t scale it up – they really want to scale it out – way out.  As importantly, they want it to be elastic. 

These are major challenges for these customers, and there is a growing sense that perhaps the architectures of the past – while the need to continue innovation in their own right, it’s time for some new options.   VMware’s acquisitions of Springsource, Gemstone (read up on the technical aspects here), use of TC server, and the key brains behind Redis are all manifestations of that new battlefront.  ‘Zilla, you do have a very good sense of what happens in the boardroom – you’ve mapped it out right GemStone- VMware goes after your Database :-)

At the same time EMC and VMware in the VCE coalition will be driving innovation in the more traditional stack with BI/DW partners (some at EMC world, some at SAP’s Sapphire).   For EMC co-opetitors in database-land, we’re of course going to support the “oldey-timey” way (massive traditional database scale) way through innovation, and simultaneously we think that this wave of change represents a flanking opportunity for us (and surely a HUGE challenge – we have big competitors here), and a big opportunity for our customers.

But while that is all true, maintaining and extending our position of storage leadership in every market segment is critical to EMC.  Storage is our cash cow – and funds our explosive expansion into adjacent markets.  “EMC=best-of-breed” is what our customers expect and demand from us.

… So without further ado, here’s what we announced and demonstrated to maintain and extend that leadership position in midrange SAN and NAS storage.

In the mid-range market, there are two overwhelming requirements:

• Simplicity
• Efficiency

Nail those, and customers like it :-)   There’s rocket science stuff, and then there’s execution stuff… (and sometimes nailing the execution stuff and making things simple and efficient require rocket science under the covers).

Today, EMC announced and demonstrated what we’re doing in the Unified Storage division to provide our customers quantum leaps in simplicity and efficiency.  

• Simplicity: Unification of the management of SAN and NAS with Unisphere & and “self service” for the VMware Administrator with the next-generation of vCenter plugins from EMC.
• Efficiency: FAST Cache, Sub-LUN FAST, and Compression, vStorage API for Array Integration – hardware offload support

Perhaps the best part?  If you’ve got a current generation EMC CLARiiON (iSCSI/FC/FCoE only) or EMC Celerra (iSCSI/FC/FCoE and NAS) platform (using the 64-bit storage engine hardware) – you will get all this in your current hardware in Q3 (just around the corner).

For the last few years, we’ve been taking these two coupled ideas of simplicity and efficiency incredibly seriously.  Behind the scenes, we’ve been beavering away, and beyond these new capabilities, huge chunks of codebases have been merging for a LONG TIME.   The CLARiiON and Celerra already merged their iSCSI target (this new FLARE release has a BIG improvement for vSphere iSCSI shops – see detail on that here), data layout/metadata logic (enabling virtual provisioning/storage pool models, next-generation snapshot use cases, auto-tiering and other things) – in case it’s not clear, the “RAID Group” as the way you provision is going bye-bye.

Today, I think we’ve materially moved the ball forward, and are delivering to our customers.   We’re not arrogant – in fact we know that there’s still much more do to on both fronts of simplicity and even more we can apply on efficiency.

BTW – we’re not stopping here.   We’re being relentless on ourselves.  We are focused on the task.   We have the resources to innovate AND execute.   I would HATE to be an EMC competitor – whether it be on the battlegrounds of storage, backup, virtualization, security, cloud.

Both topics (“What is EMC delivering around next-generation simplicity” and “What is EMC delivering around next-generation efficiency”) is long enough to warrant their own posts – so that’s what I’m going to do. 

Follow the links for details, demos, and performance/load/efficiency testing results:

• Next Generation Simplicity
  • Unisphere = we’re integrating SAN, NAS and Recoverpoint management in a “written from the ground up” Flex based UI – fast, flexible, customizable – and most importantly… simple.
  • vCenter Plugins.Next = we’re enabled complete “self-service” for VMware admins – every EMC platform, every protocol, every scale (and at the top end, customers demanded something neat)
• Next-Generation Efficiency
  • FAST Cache = we’re added the ability to have up to a 2TB read/write cache using cheap hardware.  Think 80-90% performance improvement in many cases.   Crushingly good results in View use cases.
  • Sub-LUN FAST (automated tiering) = we’re adding the ability for the array to tier at sub-LUN levels of granularity.
  • Compression = we’re adding the ability for the array to compress SAN production storage (in addition to our ability to dedupe files and compress files or filesystems in NAS) – which nets between 20-60% capacity savings (varies depending on dataset).

Awesome stuff!

Comments

[Rob Koper]

Dedupe on primary storage ? That's news ! I'm very curious about performance. This HAS TO cost some performance. I can imagine that with DD technology this cost can be reduced to hardly noticeble proportions, but still.... dedup on primary storage....

[ttgrant]

Chad,

First... I apologize for ambushing you in the hallway with pressing needs on Thursday. As usual you handled it like the ultimate professional. Don't know if I influenced the wrap up in anyway, likely _not_ as you already knew the answer. Regardless, you and Jonas wrapped up the day perfectly.

1) Simple and Efficient.
2) Don't be a doink
.... and most importantly don't go negative.

The vSpecialists rock!