Well – it’s a big day today :-)
For context – here’s the announcement, and I will also link to the recording of the 3 CEOs discussing what it is, what it isn’t, and what it means. Here’s also a summary video from John, Joe and Paul.
I’m also glad finally to be able to start talking openly – you should have seen the edits that occured to the VMworld 2009 VMware/Cisco/EMC supersession (SS5240 – which you can watch here) to tiptoe around this (if you do watch it now knowing what we’ve been working on – it’s interesting). Specifically, the inital versions of that VMworld session called out what we were doing together around Joint Engineering, Joint Solutions, and Joint Support models. Suffice it to say it go kiboshed by the secret police at all three companies – but we still managed to sneak in a lot of sneak peaks :-)
IF YOU ARE a VMware/Cisco/EMC Employee (and on your intranets) – you can get more content from http://www.vceportal.com
IF YOU ARE part of the public and interested in what we’re doing on this, there is more content here http://privatecloud.com
So, now that the cat’s out of the bag – what exactly are we talking about?
The most important thing isn’t the tech (though that is very cool, and I’ll get to that in a second), or the joint Cisco/EMC/Intel venture (though everyone always focus on that because it is evidence of joint commitment) – it’s an idea.
That idea? It’s what people who have heard people from VMware (like Paul Maritz or Steve Herrod), or Cisco (like Padmesree Warrior or Ed Bugnion) or EMC (like Joe Tucci or Chuck Hollis or myself) talk in the last year will seem very familiar:
- If you agree that x86 wins (price and price/performance – today mostly, and tomorrow almost entirely), and virtualization is the way to go – it’s possible to build infrastructure based on those assumptions in a much more standardized way, with much more cost-effective and innovative elements, and importantly - leverage VMware and VM-Aware networking, compute, and storage resources to “get out of the plumbing business”.
- This could not only saves a ton of capex dollars, just as importantly, one could get order of magnitude improvements in operational efficiency – instant self-provisioning, end-user chargeback. This could result in interesting new business and economic proposals, where you could pay-as-you-go as opposed to up front-capex centric models – if you had vendors willing to partner with you in that way.
- It requires infrastructure not only integrated with VMware, but also designed for scale-out, so you can start small, and get as big as you want (otherwise the economic models are whacky). Also – if you took the engineering resources of industry leaders and applied them in a very focused way against this use case – you could build something so integrated, it would in effect be a software mainframe built out of commodity elements.
- It also requires management that is deeply integrated across the stack – from the infrastructure up to the business application, and that spans the physical and virtual infrastructure.
- Heck – if you could do all that – you’d have standardized building blocks for people building internal clouds, and external clouds (either public or private)
That’s what we jointly announced today – execution against that idea – making the idea real.
To recap – there were 4 major parts of today’s announcement. I’m going to break this up into 4 posts – one that deals with each.
- Technology Innovations - Vblock Infrastructure Packages. Tightly integrated standardized “building blocks” Click HERE for more on this topic.
- Integrated Pre-Sales, Services and Support - Vblock Unified Customer Engagement. Engage like we’re one company, get services like we’re one company, support that is exactly like we are one company. Click HERE for more on this topic.
- Solutions Venture and Investment - Acadia. A Cisco-EMC (and Intel) joint venture to build, operate, and transfer Vblock infrastructure to organizations who want to accelerate their journey Click HERE for more on this topic.
- Partner Ecosystem Leverage - Vblock Partner Ecosystem. Click HERE for more on this topic.
There’s huge amounts of detail on each of these things – enough for twenty blog posts (more I bet!). We’ve been working on this for more than a year (we’ve been plowing away with “intense partnering” for many years, one year ago we got the “engage in joint partnership like nothing else” direction).
I’m also going to write another one where I will try to answer a question I think we will get often: Where does VMware fit in the VCE Coalition – are they still independent? The short answer is: ABSOLUTELY. Go to the link for a more detailed dialog.
It’s an exciting day – and I’m looking forward to having a dialog with you about it!
While this is an interesting announcement, there's really nothing new here. Solution providers have been putting together VMware, Cisco and EMC solutions for months. This appears, to me, to just be another marketing gimic designed to give Cisco and EMC a leg up on HP.
Posted by: Kevin | November 03, 2009 at 03:25 PM
@Kevin - respectfully, I disagree (though perhaps that's our destiny based on who I am, and your email alias).
Let me tell you a couple reasons why:
1) the management tools like EMC Ionix UIM manages the Vblock as what it is - not a set of parts, but rather a system. Likewise, EMC Ionix DCI integrates, correlates and shows dependency and root cause of the Vblock, vSphere, VMs and ultimately applications as what they are - not a set of parts, but rather a system.
2) Cisco and EMC can point to many things that make them more integrated with VMware than HPs server, network, or storage offerings (shortlist of examples: VN-link/Nexus 1000v, PowerPath/VE, vCenter pluging, storage that integrates directly with vCenter) as well as faster technology innovation that can be leveraged to be more efficient (FCoE + memory extensions on Cisco's side, scale-out storage, FAST, solid state, deduplication on EMC's side). This is the "quid pro quo" of the "best of breed" versus the "one stop shop". We're proposing a model that gets the good of both. It's up to us to execute against that, but that's what we're trying to do.
3) The ultimate express of the last two sentences of #2 is Acadia. For customers that buy into the principle that they are looking for IaaS built from best of breed, but integrated so much that they operate as an integrated offer, that ultimate expression is to ask: "can I buy this 'by the VM' but have it in my datacenter?" Acadia makes that answer "yes", whereas traditionally the only answer is "yes, if you use an external/public cloud".
Those are 3 quick reasons, at least in my book.
Thanks for the comment!
Posted by: Chad Sakac | November 03, 2009 at 04:35 PM
what is your opinion on this?
http://h18004.www1.hp.com/products/solutions/converged/main.html
sounds to me like HP says: us too
regards
Jose R
Posted by: Jose Ruelas | November 04, 2009 at 02:58 PM
@Jose - thanks for the comment. HP also has an full stack, and partners with VMware. So, what do we think is the difference?
Well - ask the following of HP (and anyone offering a VMware-centric stack):
- On storage: 1) what vStorage APIs do you leverage today?; 2) please demonstrate your VAAI support in K/L.next.; 3) what vCenter extensions do you offer for integrated management? 4) can you snapshot datastores and VMs (and the apps in VMs) in VMware and App-integrated ways?; 5) Show me your SRM solution; 6) If you're a big enterprise customer or a service providers, show how you can horizontally scale ports, brains, disks - at very large scale, starting small and getting big. 7) Do you support 10GbE/FCoE? And, what is your plan to implement and support FC-BB-6?
- On network: 1) Do you support network-level visibility to the VM level? 2) what are my options for network virtualization from the guest, to the adapters, through the network, all the way to the target/server? 3) can I manage my LAN/SAN in a single integrated fashion?
- On compute: 1) how do I manage a blade, a chassis and multiple chassis' - and can I copy and paste templates across the full stack? 2) how is your server particularly designed or suited for high-density, high power VMware-oriented configurations?
- On management: Is there one tools that can handle provisioning, change control, and remediation of the entire HP stack - from server, to chassis, to network/SAN, to storage array?
I'm not an expert on HP's offerings, so am not qualified to judge them - but they key thing I'm trying to say is that the engineering efforts of Cisco and EMC are VERY focused around this use case.
Our goal is to be best of breed, but also operate like one company. Hope that helps!
Posted by: Chad Sakac | November 08, 2009 at 12:18 PM
Seems to me your last point about public and private clouds starting to use the same building blocks is the biggest differentiator. Yes all integrators have a solution, but no, not everyone has the platform to do long range DRS of the workloads from a private VBlock to a Savvis or Terremark cloud!
I write briefly about this here on my blog. http://www.uptimesoftware.com/uptimeblog/agile/just-how-disruptive-is-cloud-to-the-technology-spectrum/
Posted by: Kenneth Cheung | November 10, 2009 at 02:09 PM
This is what HP strives to be, but will be a couple of year behind in this race.
Posted by: Brian | March 17, 2010 at 12:53 AM
this part of the blog "There are a number of ideas that fall out of this, that explain, or connect, to very consistent behavior. " It was very interasante, thanks for the information!
Posted by: louis vuitton bag | July 10, 2010 at 02:14 AM
Well, This is the "quid pro quo" of the "best of breed" versus the "one stop shop". We're proposing a model that gets the good of both. It's up to us to execute against that, but that's what we're trying to do, thanx for the wonderful share.
Posted by: Tulsa Bankruptcy | February 26, 2011 at 02:44 PM