To be able to provide accurate information, Ambiental, developed a modelling platform that takes in consideration a large number of parameters and data to deliver an accurate prediction of where a flood may cause damages.
Fine tuned algorithms and machine learning make analysis more accurate all the time but they also need efficient and dependable computing platforms to provide the number crunching power required to do the job.
Ambiental used to have a small local cluster and use universities HPC (High Performance Computing) infrastructures when available but as their success meant they needed more computing power to satisfy more contracts with larger projects they needed a reliable platform with which they can plan and execute the workloads.
“The ability to be self-sufficient when it comes to data hosting, delivery of our products and running the simulations is very important to us. This drove the need to have an in-house HPC/Cloud Computing solution that we can fully control” said Justin Butler, CEO at Ambiental Risk Analytics.
Having full control of their infrastructure means Ambiental's employees don't have to wait for external resources to become available or to upload and download large amounts of data to and from external services. This greatly improved their productivity.
Many think that the Cloud is an incredibly complex infrastructure that requires massive scale deployments and a team of dedicated specialists to run. That may be true for some but not for a modern Cloud platform like OnApp.
OnApp is already powering 30% of the Public Cloud so it can be scaled massively but you can start with a single Intel appliance and some networking equipment to satisfy the needs of an SMEs.
If your business is thinking to implement a basic VMWare infrastructure with a small SAN or to deploy an hyperconverged infrastructure (HCI) then you will find that an entry level OnApp configuration will cover your requirements, allow for unlimited scalability and massively reduce your costs.
Ambiental's Private Cloud building blocks
This, for Ambiental, is just the start so everything is in place to allow them to add more computing and storage nodes at anytime. As pictured above the initial configuration just needs an Intel appliance, a 10Gb switch and a standard 1Gb switch. That's what Cloud looks like if you haven't seen it before.
The 2U Intel chassis contains 4 compute nodes which share 24 storage drives which can be mixed to satisfy the needs of your workloads.
In this case, as Ambiental needs a lot of processing power and modest number of IOPS, each node has been configured with 36 CPU cores, 256Gb of RAM and HDD drives cached by an NVMe adapter.
Workloads are replicated to 1 or more nodes for resiliency so each node is equipped with a dual 10Gb network adapter which provide the speed required to match the NVMe cached distributed storage. Cumulus Networks is the NOS (Network Operative Systems) that has been installed on the 48 ports switch as it's easy to manage, is very reliable and allows for complex configurations once there is a need to scale out of a single rack.
As Ambiental is working with Governments around the world keeping data secure is very important and that is an additional reason that made them prefer a Private Cloud.
When there is a need to share files then NextCloud is ready for internal and external users. With NextCloud customers can upload the data that needs processing knowing that is not being stored in any Public Cloud and that it won't need to leave Ambiental's offices. As NextCloud is integrated with Kopano and Collabora Office all the features generally found in suites like Office365 are covered in Ambiental's own Private Cloud.
Some may say "Why not using Public Cloud services?"
Well, Ambiental went through due diligence and got pricing from the major Public Cloud providers.
It turned out that with the cost of renting other people's infrastructure they could buy a new Intel/OnApp infrastructure roughly every 3 months.
Now Ambiental, instead of paying all the time to rent resources, can invest the massive savings in developing more innovative and efficient algorithms that will help in saving more lives.
For more information read the case study and the white paper.