A little over a week has gone by since the end of EMC World, and all the product announcements have gotten out of the bag. So, why another article about EMC World, if there are no “big reveals” left? Because I want to make sense of all of the hype, product announcements, and strategic discussions. What do the over 40 new products mean to GreenPages’ customers—both present and future? How many of those products were just cosmetic makeovers and how many are actual game changers? Why should you, our friends and extended business family, care, and what should you care about?
I will start by saying that this EMC World really did reveal some technology-leading thoughts and products, and proved that EMC has taken the lead in major storage technology strategy. EMC has always been the 800-pound gorilla of the storage industry, but for many years was far from the front of the pack. This has changed, and in a big way. Innovation still takes place mostly in the small companies on the bleeding edge of storage (SSD, virtualization across platforms, innovative file systems), but EMC has become the leading investor in storage R&D, and it shows. While they may not be inventing the coolest and most striking new storage and hardware, their pace of development and integration of that cool stuff has exponentially increased. Time to market and product refresh cycles are picking up pace. Relationships with the people who get the products in front of you (resellers, integrators and distributors) are vastly improved and much friendlier to the commercial world we all live in (as opposed to the rarified heights of the largest enterprises). The relevance of EMC products to the virtualized datacenter is clear, and the storage engineers who ran the technical sessions and laid out all the new storage, DR, and virtualization roadmaps proved that EMC is the leading storage technology firm in the world.
What are the highlights for GreenPages’ world?
Probably the biggest technology in terms of impact, IMHO, is Isilon. This is the fastest, most scalable, easy-to-manage NAS systems ever. It can grow to the petabyte range, and there is no downtime or forklift upgrades. It is “scale-out” storage, meaning you add nodes that contain processing (CPU), RAM for Cache and additional bandwidth, along with capacity in three flavors (SSD, 15K and 7.2K). This is the system of choice for any healthcare PACs application or Life Sciences data storage. It is a fantastic general-purpose NAS system as well. Isilon is the system of choice for anyone managing Big Data (large amounts of unstructured data). The entry point for this system is around 10 TB, so you don’t have to be a large company to find the value here. Isilon also has the advantage of being a true scale-out system. Some technical nuggets around Isilon OneFS Upgrade: 90% greater throughput, or 740 GB/sec; roles-based admin – SEC 17a-4 compliance; better caching (50% reduction in latency of IO intensive apps; VMware Integration: VAAI (vStorage APIs for Array Integration) and VASA (vStorage APIs for Storage Awareness).
If you are going to jump up into the big time storage array arena, the new VMAX line is arguably the one to get, for power, performance and integration with the virtualized datacenter. It has expanded to the VMAX 10, 20 (current), and 40. The top of the line sports 8 controllers, scales up to 4 PB, has up to 32 2.8 GHz Xeon 6-core processors, 1 TB usable RAM, 2.5” drives, and uses MLC SSD drives (bringing that cost of the flash drive down into the lower atmosphere). The latest development of the auto-tiering software FAST allows IBM and HDS storage to be a “tier” of storage for the VMAX. Other arrays will be added soon.
VNXe 3150 storage system offers up to 50% more performance and capacity in an entry level system. This system includes 10 GbE connectivity, Solid State Storage and the first production storage system (that I have heard of) that uses the latest Intel CPU, Sandy Bridge. Who says EMC product lifecycles are slow and behind the times??
VPLEX Metro/VPLEX Geo solutions have some significant upgrades, including integration with RecoverPoint and SRM, more performance and scalability; and Oracle RAC up to 100 KM apart. If you want to federate your datacenters, introduce “stretch clusters” and have both an HA and DR strategy, this is the industry leader now.
The VNX Series has more than a few improvements: lower price SSDs, RAID types that can be mixed in FAST; 256 snaps per LUN; connector for vCOPs; EMC Storage Analytics Suite based on vCOPs; AppSync to replace/improve Replication Manager.
The new VSPEX Proven Infrastructure includes EMC's VNX and VNXe hybrid storage arrays, along with Avamar software and Data Domain backup appliances. The cloud platform also includes processors from Intel, switches from Brocade, servers from Cisco, and software from Citrix, Microsoft HyperV and VMware. Avamar and Data Domain products will offer data deduplication to users, while EMC's Fully Automated Storage Tiering (FAST), will offer data migration between varying disk storage arrays based on data use patterns. There are initially 14 VSPEX configurations, which EMC said represent the most popular use cases for companies moving to cloud computing.
Data Domain & Avamar upgrades include the DD990 with an Intel Sandy Bridge CPU, doubling the performance of the DD890 – 28 PB, 16 TB/hr throughput; tight integration of Avamar with VMware, including Hyper-V, SAP, Sybase, SQL2012 – recovery is 30 times faster than NBU/V-Ray.
Vfcache PCIe NAND Flash Card is a server side I/O enhancement that pushes Flash Cache to the server, but integrates Cache management with the VNX array FAST Cache. This will prove to be a huge deal for mission critical applications running on VMware, since I/O will no longer be a bottleneck even for the most demanding applications. Combine this with Sandy Bridge CPUs and the UCS system with the latest M3 servers and you will have the world’s most powerful server virtualization platform!
DataBridge is a “mash-up” of nearly any storage or system management tool into a common pane of glass, not intended to be a discovery or management tool but, rather, to be a place where all of the discovery tools can deliver their data to one place. This combines EMC infrastructure data sources along with non-EMC data sources with business logic from customer organizations. Stay tuned for more on this.
There are lots of other deep technical messages that were talked about in the sessions that ran for three solid days, not counting the unbelievable Lab sessions. Those Lab sessions are now available for demo purposes. You can see any EMC technology from implementation to configuration just by contacting GreenPages and asking for your Friendly Neighborhood Storage Guy!!
One final thought I would like to stress: efficiency. EMC is sending a smart business message of efficiency, using VNX as example. Storage is far outstripping storage advances and IT budgets. All is not hopeless, however. You can improve efficiency with dedupe/compression, auto tiering; Flash allows storage to keep up with Moore's Law; you can consolidate file servers with virtual file servers (we have done this with many GreenPages customers when consolidating servers in VMware). Files are the main culprit. How will you manage it, quotas or content management? What will you chose? How will you manage your data without the money or work force you think you might need?