VMware stays the course - executes - progresses on fight for long term relevance
VMworlds are always overloaded with new information and announcement, remarkably VMware only released 4 press releases (find them all here) – a welcome change to the floods of the past – but each press releases was like half a dozen normal press releases. So tough to pick my top 3 takeaways – but here you go:
Live VM transfer in the SDDC world - As part of the Day #1 keynote we witnessed a world premiere, and a remarkable technical feat – the live transfer of a VM from one datacenter (or cloud if you will) to another one. Very remarkable and of huge value for VMware customers and the vendor itself – as it gets VM based loads portable, which opens a lot of use cases. VMware itself was not shy to push for HA / DR plans, interesting how all early players to public cloud stress the same use cases – it was the same for Microsoft and IBM – only respectively two and one years ago. It is good to see that VMware has delivered this capability, though for now it’s in tech preview and I did not catch a go live date. But given the importance I am sure we will hear and learn as soon as possible when this capability is generally available. VMware CEO Gelsinger was right to point out that no other cloud provider has been able to demo the same.
On the vCloud Air side, VMware makes progress, too – storage partnerships with EMC (no surprise) and Google (more surprising) are good moves. A database offering for Microsoft SQL Server (more are supposed to come) are another indication that VMware understands it needs to create opportunities for customers to create load on its public cloud. But the road is long here. Adding the capability to run EMC Federation sister Pivotal’s PaaS product CloudFoundry is a key move along the same lines.
And overall VMware makes progress on its SDDC vision from last year, a lot of work on the software defined networking side has been done, it can be combined with software defined storage and of course run well on the VMware Evo hyper converged appliance. Security was stressed at many points and in the Day #2 keynote we even saw a network controlled data access to a GRC application, an interesting perspective and approach to a problem that traditionally is solved as an application security problem.
But make no mistake, the success of these initiatives and products will determine how long enterprise will keep running on premise servers, so they are more than strategic for VMware.
Photon as the path forward – VMware also realized that it needs to play with many of the open source technologies that are out there, always with the view of security and to protect existing investments into VMware products. The whole Photon family is nothing else than that, with a VMware offered Linux at its core, that runs of course – on a VM. Thanks to the popularity of Linux, almost all open source products become available, the most prominent ones are micro services. In the process of that VMware needs to make some bold assumptions (or in new Silicon Valley speak I learnt this week – opinions) – e.g. one VM for one container, but VMware re-assured us that the VM footprint is small and the provisioning is very fast. We will have to see what works faster and more cost effective – running containers natively vs running them through VMware VMs and admin tools. What is clear is that VMware offers a way to run these technologies in a safe and well understood way for IT departments. And as such there is a lot of value for the existing customers in what VMware does with Photon.
EUC gets a boost from Microsoft – The EUC group is making good progress on its mission to create additional growth for VMware in an adjacent software category. It’s good to see rationalization that is to be expected like ‘Project A2’ (think AirWatch meets AppVolumes) and Horizon running on Evo farms. VMware using more VMware is something that is good for customer and likely something investors want to see, too. Project Enzo appears to be on track and should give VMware more addressable market by lowering the cost of running virtual desktops. Every year I come to VMworld I ask if the virtual desktop bubble is going to come to a pop – and every year we come closer, but it has not popped. In a ‘and now hell freezes over’ VMware had a Microsoft executive in Poonen’s keynote – thanks to Windows 10 being the same code / platform across PCs, tablets and smart phones, AirWatch has become more than a MDM tool, and can now also manage PCs with Windows 10, making both vendors best friends. Sharing progress of a large university that has virtualized 60k student desktops and plans to go to 70k is encouraging for the division prospects
It is good when vendors stay the course, as software is like wine, it takes time to become good. So no disappointment by this analyst that VMware showed execution along the same strategy. The gambit for VMware is, if it can grow new products fast enough as on premise compute virtualization dripples away to the cloud. When I asked CEO Gelsinger on this, he shared the high level VMware equation – with impressive openness: VMware plans to make up the losses from reduced compute virtualization on premise by having widened it addressable market with SDDC (add networking, add storage going from 6B to 30B US$), revenue growth in its EUC portfolio and increased vCloud air revenues. We also learnt that VMware services is a 500M US$ business, growing at 20% YoY. An interesting equation with 4 variables, the speed of servers moving to the public cloud being the key variable – but as no one knows how fast that is happening, we can only plug in the numbers in the VMware ‘formula’.
On the public cloud side I remain concerned that VMware is not pushing vCloudAir harder. CIOs should get emails (or other solicitation) along the lines that VMware knows that load A or B can be run better in … e.g. vCloudAir. I am still not sure what is holding VMware back – capital, business practice, partner relationships, product capability, it remains an area to watch. What we know from VMware customers is that the partner approach to public cloud is not working, contracts, costs and SLAs need to be negotiated separately across geographic regions (as there is no global partner) and that is less attractive than negotiating with the public cloud vendors. But no vendor knows better what enterprises run, leveraging that know how will require (IMHO) more VMware data centers. The ‘capital light’ approach that VMware has been charting the last 12 months is not going succeed. A year ago VMware talked about datacenter rollout and location, this year it was mum on the topic, not encouraging. Thinking in-house load – EUC should generate some substantial load for VMware at some point, so seeing the synergies of running on EVO, Linux with Enzo are all good news. But it cannot be enough load to compete on the scale of the public cloud competitors.
But overall a good VMworld for VMware customers – VMware is doing what they said they will do – key technical breakthroughs like moving VMs are looking like they are getting ready for production soon, SDDC keeps evolving, Project Enzo appears to be on track, new opportunities arise from Photon and Project A2 – so there is a lot that VMware customers need to get their arms around and digest. We will be there to explain, de-cipher and analyze.
More on VMWare
- Musings – What will it be this year at VMWorld - read here
- Market Move - EMC to acquire Virtustream - More paths to the cloud - read here
- News Analysis - VMware makes progress towards the (hybrid) cloud - now let's watch the adoption - read here
- Speed Briefings at VMworld - read here
- Event Report - VMware makes a lot of progress - but the holy grail is still missing - read here.
- First Take - VMware's VMworld Day 1 Keynote - Top 3 Takeaways - read here.
- Progress Report - Good start for VMware EUC - time for 2nd inning - read here.
- Speed Briefings at VMworld - inside and outside the VMware ecosystem - read here.
- VMware defies conventional destiny - SDDC to the rescue - read here