AllightSykes primarily services the mining industry, fluctuating resource prices meant the organisation needed the ability to scale its operations in parallel with the rise and fall in mining investment.
- Fluctuating resource prices required the ability to scale operations on-demand
- Twenty-hour data extraction times threatened to derail ERP cloud migration
- Frequent timeouts made cloud migration unfeasible
- Prohibitive costs of adding network capacity mandated a different solution
- Riverbed SteelHead CX for Amazon Web Services (AWS)
- Ability to scale operations in parallel with the rise and fall in mining investment
- Successful ERP migration to the cloud by optimising data extraction from AWS
- Cloud flexibility without sacrificing performance
- Cost-savings resulting from eliminating the need to expand WAN links
AllightSykes (http://www.allightsykes.com/) is the industry leader in mobile lighting towers and auto-prime dewatering solutions, in addition to being the Australian dealers for Perkins engines, FG Wilson Power generation, Sulzer and Dragflow pumps as well as SMC lighting towers.
With manufacturing facilities in WA, NSW, Dubai and USA, AllightSykes is an international partner of choice for mining, construction, civil and hire businesses
offering first class products and services to a dynamic and diverse market.
Challenge: ERP cloud migration threatened by data extraction delays
WA’s AllightSykes primarily services the mining industry, fluctuating resource prices meant the organisation needed the ability to scale its operations in parallel with the rise and fall in mining investment—an activity inexorably linked to market activity beyond its control.
In order to achieve this, it decided to transition its on-premises ERP application to Amazon Web Services (AWS), which could provide the necessary on-demand
scalability. It quickly became apparent, however, that network traffic would need to be optimised for the migration to be viable.
With a requirement to refresh transactional data regularly it was decided to keep the company’s Business Performance Warehouse (BPW) server on-premises so that reports could be refreshed efficiently to give the business up-to-date insights into opportunities, improve processes and enable productivity gains.
“During our initial testing, the nightly data extraction jobs were never really successful,” said Chris Uusimaki, ICT Manager at AllightSykes. “Even though for the most part it was the same data being replicated—supplier lists and items lists for example—it would take upwards of 20 hours to extract and would often time out.”
Solution: Riverbed SteelHead CX for Amazon Web Services (AWS)
Uusimaki considered increasing the bandwidth of WAN links, but discarded this option quickly as the links would need to be increased to 100 megabits per second or more, which was cost-prohibitive.
As the previous ERP system was hosted on-site, AllightSykes had little ability to ramp up or ramp down its operations in line with demand. While this was not an issue during Australia’s mining boom between 2005 and 2012, falling resource prices and a subsequent fall in mining investment, meant the organisation could not afford to run services at full capacity and bear the infrastructure costs
involved in doing so.
While the flexibility and cost savings offered through the public cloud were identified as key drivers for AllightSykes’ future plans, Uusimaki had to overcome the same latency and bandwidth constraints that often challenge physical
infrastructure environments—only at a greater scale given the added geographical distance between user and data inherent in cloud delivery.
AllightSykes was already using Riverbed SteelHead appliances to optimise network traffic and accelerate application performance across six Australian sites,
so Uusimaki decided to investigate Riverbed’s solution for optimising the performance of cloud and SaaS applications across hybrid WANs: Riverbed SteelHead CX for Cloud.
“During our initial testing phase, we discovered that data ex tractions from our cloud based ERP platform to our on-premise BPW server just wasn’t feasible without optimisation or significant increases in link capacity. We tested the Riverbed solution with just a trial license first and the proof was in the pudding—it was significantly faster."
Benefits: Optimisation lets ERP migration take flight
“To put it simply, maintaining our BPW server on-premise was not going to work in its current form without Riverbed. We’ve been able to take advantage of the scalability and cost savings of cloud, without having to sacrifice performance as
a result,” said Uusimaki.
“Prior to Riverbed optimisation, hourly extracts would take up to five hours to complete. With Riverbed, we’ve been able to reduce this to about 45 minutes.”
The nightly extracts, which were so problematic for AllightSykes, have been cut from 20 hours to just six.
Uusimaki said the scalability of migrating to the cloud has provided enormous benefits to the business, particularly given that demand in AllightSykes’ industry is driven by fluctuating resource prices.
“Cost has traditionally been the primary motivation to move to the cloud but, for us at least, flexibility and scalability have been the key benefits. It is very difficult for us to capacity plan in our industry, where it’s tough to see if the next six
months will be booming as miners capitalise on resource prices. In the cloud environment, it’s no longer an issue.”
According to Uusimaki, not only has the ERP migration been a success, but he received the best kind of feedback from his end-users he could’ve imagined—none.
“With the optimisation, the cloud migration project has been so seamless that it hasn’t raised a single eyebrow. End-users haven’t noticed any difference in performance compared to when ERP was on-premises, and that is exactly what I want to hear.”