It seems to be the season for tech company announcements. I was recently briefed by Oracle on their Ravello on Oracle Cloud Infrastructure announcement and thought I’d take the time to provide some coverage.
What’s a Ravello?
“Ravello is an overlay cloud that enables enterprises to run their VMware and KVM workloads with DC-like (L2) networking ‘as-is’ on public cloud without any modifications”. It’s pretty cool stuff, and I’ve covered it briefly in the past. They’ve been around for a while and were acquired by Oracle last year. The held a briefing day for bloggers in early 2017, and Chris Wahl did a comprehensive write-up here.
HVX
The technology components are a:
- High-performance nested virtualisation engine (or nested hypervisor);
- Software-defined network; and
- Storage overlay.
[image courtesy of Oracle]
The management layer manages the technology components, provides the user interface and API for all environment definitions and deployments and handles image management and monitoring. Ravello in its current iteration is software-based, nested virtualisation. This is what you may have used in the past to run ESXi on AWS or GCP.
[image courtesy of Oracle]
Ravello on Oracle Cloud Infrastructure
Ravello on Oracle Cloud Infrastructure (OCI) provides you with the option of leveraging either “hardware-assisted, nested virtualisation” or bare-metal.
[images courtesy of Oracle]
Oracle are excited about the potential performance gains from running Ravello on OCI, stating that there is up to a 14x performance improvement over running Ravello on other cloud services. The key here is that they’ve developed extensions that integrate directly with Oracle’s Cloud platform. Makes sense when you consider they purchased Ravello for reasons.
Why Would You?
So why would you use Ravello? It provides enterprises with the ability to “take any VMware based multi-VM application and run it on public cloud without making any changes”. You don’t have to worry about:
- Re-platforming – You normally can’t run VMware VMs on public clouds.
- P2V Conversions – Your physical hosts can’t go to the public cloud.
- Re-networking – Layer 2? Nope.
- Re-configuration – What about all of your networking and security appliances?
This is all hard to do and points to the need to re-write your applications and re-architect your platforms. Sounds expensive and time-consuming and there are other things people would rather be doing.
Conclusion and Further Reading
I am absolutely an advocate for architecting applications to run natively on cloud infrastructure. I don’t think that lift and shift is a sustainable approach to cloud adoption by any stretch. That said, I’ve worked in plenty of large enterprises running applications that are poorly understood and nonetheless critical to the business. Yes, it’s silly. But if you’ve spent any time in any enterprise you’ll start to realise that silly is quite a common modus operandi. Coupled with increasing pressure on CxOs to reduce their on-premises footprint and you’ll see that this technology is something of a life vest for enterprises struggling to make the leap from on-premises to public cloud with minimal modification to their existing applications.
I don’t know what this service will cost you, so I can’t tell you whether this service will provide you with value for money. That’s something you’re better off speaking to Oracle about. Sometimes return on investment is hard to judge unless you’re against the wall with no alternatives. I’ll always say you should re-write your apps rather than lift and shift, but sometimes you don’t have the choice. If you’re in that position, you should consider Ravello’s offering. You can sign up for a free trial here. You can read Oracle’s post on the news here, and Tim’s insights here.