Cloud Compute Automates Success

A Successful Cloud Optimisation Project for SLR

The Overview

With a workforce of over 2000 people, across 100+ offices and clients in 6 regions, SLR Consulting (SLR) needed a way to manage their client’s data and perform data processing in a scalable and secure way.

The Customer

SLR Consulting is a leader in environmental and advisory solutions: helping clients achieve their sustainability goals. SLR’s stated ambition is to be recognised as the global leader in environmental and advisory solutions, helping their clients to achieve their sustainability goals.  Cloud technology is playing a leading role in this sector, with increased use of data and computational models aiding environmental decision making.

The Problem

SLR Consulting work with their clients on large scale groundwater models that require considerable amounts of compute power, and often take weeks to run. These models are critically important to the success of their consulting engagements and need to be delivered in a timely manner.

In their current environment the SLR engineers were using a network of workstations that sets a limit on the size and speed of processing, they were also tricky and time consuming to manage. Additionally, the workstations were often being used for other purposes, leading to congestion of processing power and often the need schedule or even re-run calculations.

The current processing speed for large scale models was 3 to 4 weeks. SLR needed a new, secure and predictable way to run large scale modelling quickly and cost-effectively.

I love helping our clients like this, the impact and benefits to users and business are so obvious! The Azure Cloud gives IT professionals the opportunity to have a real impact on the SLR business model, on profitability and on freeing up valuable consulting time – we simply couldn’t have done this without Azure Cloud.

The Solution

cubesys was engaged to help SLR adopt and optimise Azure usage. During this process, the above problem surfaced and also came with a tight project deadline, for a large client. The modelling run they needed was looking like it wouldn’t meet the project deadline. This was causing delays and had potential of damaging SLR’s excellent reputation.

cubesys cloud automation engineers stepped up to the challenge and worked closely with the SLR Consulting team to adopt a new, cloud-based approach.

Using a Microsoft cloud service called ‘Azure Batch’, the team created a new agile approach, which allows SLR to run large-scale, parallel and high-performing computing batch jobs efficiently. Azure Batch creates and manages a pool of on-demand virtual machines that allows SLR to install the modelling applications, schedule jobs and reduce processing time from 3 to 4 weeks down to 4 hours.

This solution provides SLR the ability to scale from 1 to thousands of virtual machines in an instant.  Its available on-demand, its fast, secure, and cost effective – the solution is turned on and paid for when needed.

The Results

SLR Consulting no longer need to spend valuable consultant time managing a network of workstations, nor weeks of waiting to provide their clients with the results of modelling calculations. The Cloud-based solution is automated, secure, fast, and cost-effective. Agile solutions like these provide the SLR team a competitive edge and allows them to focus on the clients’ needs, knowing they can count on timely data from computational modelling runs.

And the original project?  It used Azure Batch to spin up 175 virtual machines for a few hours, completed over 1000 model runs, and got the results needed early and avoided project delays. We delivered a successful project for a satisfied client.