Feeds

Converged IT stacks threaten storage interfaces

Vertical optimisation could oust layered standard interfaces

Beginner's guide to SSL certificates

Comment In a world of converged IT stacks the standard interfaces on which external storage depends are threatened.

External storage depends upon standard interfaces: Ethernet and NFS/CIFS/TCP/IP for filers and iSCSI and Fibre Channel for block access. We have a layered approach to connecting servers and external storage, and also servers and networking resources. This layering supports and encourages multi-vendor creativity within the layers but limits optimisation. That is made easier if you have a vertically integrated stack.

What NetApp can do with one stack - Data Ontap and its commodity-based hardware - several other vendors do with multiple products. HDS, EMC, Dell etc have separate block and file access products, such as EMC's CLARRiON and Celerra lines.

The two main IT stack convergers, Cisco and HP, are competing with each other, but are also up against the non-converged IT stack vendors, who push their best-of breed products at each layer of the stack.

HP has the most integrated and converged IT stack offering in its in-house product set: ProLiant servers integrated with ProCurve networking and StorageWorks drive arrays. Soon all the drive arrays will likely be HP servers running array control software. Cisco has had to cut a deal with EMC to get its storage, and the integration possibilities there could well be limited by EMC's need to continue selling the bulk of its storage products to connect to multi-vendor non-converged server+networking+storage IT Stacks.

Not so HP.

It can start integrating servers, storage and networking far more closely because it owns technologies in all three areas, and because a large proportion of the sales of each technology area are to HP's own customers. If HP can optimise the efficiency, performance and costs of its Matrix converged IT stack so it offers clear benefits to HP customers that are not possible with the Cisco California stack, or with the non-converged stacks, then HP will hoover up lots and lots of enterprise IT data centre dollars.

Cisco will find itself forced to answer this, and so will the two other main system/server vendors, Dell and IBM, and lesser suppliers such as Fujitsu. If they perceive that optimising vertically offers a greater pay-off than continuing to optimise across separated layers then that's what they will do.

In fact IBM is already doing some of this to an extent. Its VDS, Virtual Disk System, is an integration of DS300 and DS5000 array and its SAN Volume Controller, which itself is an IBM X server running IBM's own software. Its SONAS product is also a combination aka integration of IBM's own technologies. Suppose VDS was further integrated and included as part of a bladed IBM server offering?

HP and the others could take the view that, when you're mixing and combining bladed and virtualised servers, storage and networking in a single rack, then using multiple storage networking protocols and connection media meant for LAN and campus distances is a waste of resources. Why have both iSCSI and Fibre Channel? It's all Ethernet. Why have both iSCSI and FCoE? It's all block access and data centre Ethernet does not lose packets, so we can cut TCP/IP out of the equation and get rid of iSCSI. Why waste server cycles - or adapter cycles - wrapping data in networking protocol wrappers that there really isn't a need for?

Direct-attach storage is used for both block and file access across a single wire. Look at a Matrix or a California rack and it's easy to make the jump to thinking that all the storage in it is basically direct-attach storage. So directly attach it and share it with an extended PCIe fabric - all the servers have PCIe anyway.

Another advantage that would incline converged IT stack vendors to optimise vertically is that doing so could be the thin end of a very satisfactory wedge. If they integrate and optimise vertically and thereby get specific advantages then the competitor attach rate in their accounts will fall. Would HP like a storage optimisation strategy that has the additional benefit of decreasing the EMC and NetApp attach rate in its accounts? Of course it would. Ditto Cisco/EMC.

I'm not saying this will happen. But there will be those within the engineering and strategy shops of converged IT stack vendors that argue that optimising vertically will give the vendor more benefit bangs for each optimising buck, and that it will cut the competitor attach rate. What's not to like about either?

Of course vertical optimising supporters will say they support the standard interfaces, but they'll also say that if you use this special technology they've developed then you'll use 10 per cent less electricity per year, need 10 per cent fewer servers, and can run 10 per cent more virtual machines. Wouldn't you like to cut your annual total cost of ownership by 30 per cent? Can't you just see this sort of messaging coming out?

If this happens then stand-alone networking, storage and server vendors will be out in the cold, unable to sell to data centres moving to converged IT stacks. None of this is going to happen for several years, but we at El Reg reckon it probably will. ®

Beginner's guide to SSL certificates

More from The Register

next story
The cloud that goes puff: Seagate Central home NAS woes
4TB of home storage is great, until you wake up to a dead device
Azure TITSUP caused by INFINITE LOOP
Fat fingered geo-block kept Aussies in the dark
You think the CLOUD's insecure? It's BETTER than UK.GOV's DATA CENTRES
We don't even know where some of them ARE – Maude
Intel offers ingenious piece of 10TB 3D NAND chippery
The race for next generation flash capacity now on
Want to STUFF Facebook with blatant ADVERTISING? Fine! But you must PAY
Pony up or push off, Zuck tells social marketeers
Oi, Europe! Tell US feds to GTFO of our servers, say Microsoft and pals
By writing a really angry letter about how it's harming our cloud business, ta
Astro-boffins start opening universe simulation data
Got a supercomputer? Want to simulate a universe? Here you go
prev story

Whitepapers

Why cloud backup?
Combining the latest advancements in disk-based backup with secure, integrated, cloud technologies offer organizations fast and assured recovery of their critical enterprise data.
Forging a new future with identity relationship management
Learn about ForgeRock's next generation IRM platform and how it is designed to empower CEOS's and enterprises to engage with consumers.
Reg Reader Research: SaaS based Email and Office Productivity Tools
Read this Reg reader report which provides advice and guidance for SMBs towards the use of SaaS based email and Office productivity tools.
Getting ahead of the compliance curve
Learn about new services that make it easy to discover and manage certificates across the enterprise and how to get ahead of the compliance curve.
Storage capacity and performance optimization at Mizuno USA
Mizuno USA turn to Tegile storage technology to solve both their SAN and backup issues.