Virtualised Fruit

Yet another fruity blog taking a byte into virtualisation

Ravello 2017 Bloggers Conference – Opening Post #RBD1

Firstly I just wanted to open this up and thank John, Kat, Amy, Oracle and Ravello for selecting me as a delegate for this event. I thoroughly enjoyed it and I certainly hope we can stay engaged and make it to the next one. I certainly hope this is many more events to come.

The main point I want to make even if you don’t read all this and subsequent posts is that wow Oracle is back in the game. I know like many I did not know what to expect out of the day but I am glad about the way Oracle/Ravello handled the day and it’s certainly put Oracle back on my roadmap. I feel they are really bringing something to the cloud table and yes this is not a one size fits all but I can see why they have chosen to do things the way they have. For me being able to have the Oracle support and the cost/flexibility of running up a baremetal machine like a cloud server is great!

I also love they fact as they put it this was a good, bad and/or ugly warts and all session of which they wanted our feedback. It was great to see a vendor willing to engage with the community and even offer us a chance to say no that’s not a good idea and take it on board as you could clearly see they wanted to be transparent to allow the product to grow in the right direction.

The most powerful message for me of the day to take away was that what you have to change is nothing

Screen Shot 2017-05-24 at 11.04.59.png

We all arrived at Oracle’s conference center just after 08:00 and by this time my body was slowly starting to catch up

The day was kicked off by a good breakfast and we were provided time to network around our table and get to know one another.

John provided a great opening speech as always

At this point he handed the baton over to Navin Thadani and he at long last let us know why the industry felt they had gone quiet. Their first objective was to tackle integration the 2nd was product development. I felt this helped put all our minds at rest and set the tone for the day.

Clay Magouyrk was then up and started to cover off Oracle’s cloud vision and how they are looking to execute this. He covered off their background and how Software as a service is not new to them, they have been at this for 10+ years. The main driver for all this was customer integrations and customers wanting to modify the software. Eventually, restrictions started to stop the users flexibility so things needed to change.

Because of this Oracle took a step back and started up Clay’s team as they wanted to go into this for the long game. This is where the 2nd generation cloud came in and they started to work on the limitations of each of the already existing systems in their portfolio. Everything was started from the ground up even down to people, systems, infrastructure and networking.

What this drove Oracle to do is to try and retain the best bits such as self-service, pay for what you use with it being elastic but they need to balance this with the enterprise requirements such as security, pricing, performance and SLAs

OracleNGCloud_Page_02.jpg

Here come the acronyms now though…. There is now High Available zones and Domains. Their spin on zones is a DC within a region of which I feel is a much easier way of understanding this.

OracleNGCloud_Page_03.jpgOracleNGCloud_Page_04.jpg

The network in their instance is that it’s not oversubscribed along with off box networking virtualisation to allow the customer to manage theses aspects. The nice thing here they noted is none of this management is running on the compute nodes thus it will not affect any performance.

OracleNGCloud_Page_05.jpgOracleNGCloud_Page_06.jpg

The key thing this allows is that them to run is baremetals systems so it’s possible to have full root access to the compute node (apart from ILOM) to integrate into the rest of your stack. It then also allows intergrations with other devices such as F5s and other nodes that can be plugged in but still managed.

OracleNGCloud_Page_08

They also support direct connections with cross connects and VPN to gain access to these networks along with load balancing.

OracleNGCloud_Page_10.jpgOracleNGCloud_Page_11.jpg

I dont have pricing to hand but I know transit is by far one of the most expensive components of the Cloud of which ties you know. This is one thing they want to change so that you have choice. They have shown us some very attractive offerings and once these are published I will share them as they are a breath of fresh air in our industry.

To ensure all this all works there are redundant links between that data centres of which work in opposite paths. This does mean the minimum buy-in is always two ie 2 VPN or MPLS connections into each availability domain. The key thing is this needs to scale and no matter what a DC can always lose connectivity.

The Dyn DNS was a shock to me. I didn’t know they had bought them but this will also start to integrate throughout the year to even allow this withinside the network. This could help with DC failovers and public IPs

OracleNGCloud_Page_12.jpg

I did certainly like Eric’s marketing advice for this slide

 

As expected this offering is the best in breed for Oracle databases in the cloud

The next slide showed us why we were all there, the Ravello integration.  My only issue with this slide and it says great for X but again it can also be great for Production usage with the right architecture. I feel they really missed this punch line but didnt expand and offer this message throughour the rest of the day.

OracleNGCloud_Page_15.jpg

To follow up from this we started to cover the cloud table stakes. I think this summarizes many organisations plus and pain points really well.

OracleNGCloud_Page_17.jpg

Next Clay highlighted the main points, Uncompromised security such as covering physical isolation, high-security network, segregated zones and audits along with governance and control ensuring separate production/test/dev and separate IT and Dev responsibilities.

The next part was quite exciting for me which is the bare metal solution. In essence, they don’t care what you run on here it could be VMware, Microsoft, Citrix etc but for me, this is a great idea. At the moment it’s great putting apps in the cloud but this way you could host VDIs close to the app to prevent data ingress charges or even old CRM or Backoffice packages to reduce your latency but it takes away they headache of managing a data centre and the other bits that go with it.

OracleNGCloud_Page_20.jpg

At this stage I did ask about  a London DC ws on their roadmap somewhat selfishly but I have been advised they hope this is coming very late in 2017 to expand on from Frankfurt.

The slides then went off with some NDA stuff of which when I can share I will but it all very excting things.

At this point I am going to stop this post as I feel the other topics deserve posts of their own with what is coming up, a deep dive on HVX and the Ravello roadmap as it currently stands. But overall the day was great and I cannot wait for the next one and what they team will share with us over the coming weeks so watch this space.

As always I must put these view are my own and may not necessarily be that of my employer or hosts of the event. The following was intended to outline Oracles general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle.

RBD1

Gareth • May 24, 2017


Previous Post

Leave a Reply

Your email address will not be published / Required fields are marked *