Fresh produce distributor Minor, Weir and Willis (MWW) has moved to set up a private VSPEX converged infrastructure to support IT across its growing UK operations, as part of a major overhaul of its systems.
MWW, which had turnover of over £100 million last year, provides fresh produce to a range of supermarkets in the UK, including Tesco, Asda and Morrison’s, as well as a contract with food service firm Compass, and a business delivering to schools as part of a government scheme.
The firm’s main production operations are based in Birmingham and services over 100 IT using staff across its entire business which totals approximately 450 employees. The main business also consists of subsidiary companies, with separate premises in other regions of the UK including Evesham, Peterborough and Chesham, with each office currently running their own server and software in silos, linked by an MPLS.
In early 2012 the company started to investigate the possibility of upgrading its infrastructure in order to facilitate the introduction of a new ERP system, Microsoft Dynamics NAV 2013, which would be used to provide financials, reporting and supply chain management for its entire UK business.
The company had previously used a number of legacy applications including Sage Line 100 to support its operations, and was heavily reliant on Excel. To support its applications, MWW was previously running four physical HP servers. This included two HP ML350 twin quad core processors with 16GB RAM in an active-passive configuration, mirrored across independent storage systems, as well as two HP ML310 servers, used as domain controllers.
However the company experienced problems with a number of outages affecting staff productivity, and decided to refresh a number of applications, particularly its ERP system, which necessitated a move to a new hardware setup.
“We needed to have the right platform to put it all on, as it wouldn’t run on the old one,” said Rajinder Gill, chief financial officer at MWW. “We were having problems putting virtual machines on the existing servers – we were at capacity. NAV is hungry in terms of needing more computing power with SQL, so we had to put the infrastructure in place.”
In addition the company has been looking to expand its operations, with a move to a new production facility being considered. This led to demands for IT infrastructure that could easily scale as the company itself grew, and part of the business need centred around setting up a separate disaster recovery environment to enable the agility required to facilitate such a move.
“We have been looking for a new facility for a couple of years, and that was one of the design characteristics of building the server and SAN estate," Gill explained. "It was a case of going for a production and DR zone. This meant that if we do move to new premises, we can shut down the production end, run on the DR zone, before bringing the production end live on the new site and power up there. We could then shut down the DR site and complete the move. That was one of the main considerations.”
Gill began speaking to “box-shifter” distributors such as Misco and Insight as part of a three month investigation into the introduction of new infrastructure, with HP initially at the top of a list of vendors to provide new servers and SANs. However a discussion with Cisco and EMC reseller partner Synapse 360, led to the decision to implement the VSPEX package, encompasing EMC and Cisco hardware with VMware virtualisation, as an off-the-shelf solution.
The reseller agreed to do all the work around the provision of a fully integrated system, and a contract was signed off by the MWW board within a week, by the end of June 2012. Two EMC SANs and the Cisco Unified Computing System (UCS) servers arrived within a few weeks, and Synapse 360 had the system configured and operational by August.
“If I had known about the likes of Synapse and VSPEX in the first place, and realised the amount of pain it saved, I would have gone to them straight away,” Gill said.
The company briefly looked at implementing a Dell solution, but were put off by past experiences amongst IT staff, and considered the vendor to be more suited to providing desktops in other parts of its business, with question marks over the “poor build quality” of the vendor’s server hardware..
“Dell we discounted because my IT department have an opinion on their status in the market regarding quality: they are brilliant for desktops and laptops, but not necessarily for back office infrastructure.”
“We wanted something that was more enterprise based, where Cisco and EMC are leaders. We wanted the best of the best rather than what we consider ‘middle of the road’, or not enterprise level. In fact we were wary of the Cisco and EMC types because we thought we couldn’t afford them.”
Gill was surprised to find that the VSPEX system came in under the budget for the procurement, while a comparative HP setup with lower spec was approximately 20 percent over the budget when bought through distribution.
MWW is now running seven UCS servers with 2GHz twin eight core processors and 64GB of RAM, with the capacity to run up to 100 VMs. This is in addition to the remaining four HP servers, and two new HP servers for replication, making a total of 13 currently in use, alongside two EMC SANs.
At the same time as the hardware overhaul in June last year, MWW has been introducing new software applications, including licence refreshes for its Office suite and Sharepoint 2010, running on the firm’s intranet for collaboration. There is also an Exchange email server file print systems, legacy systems such as the SAGE Line 100 financial system, and SQL server for the backend databases.
The Dynamics NAV implementation is also now up and running, Gill said, and will allow the IT department to finally switch off Sage.
Gill said that one of the main operational benefits of the new hardware setup has been a significant reduction in down time. While there were weekly outages using Sage software on the old infrastructure, there have been no outages since implementation, and planned outages have also been masively reduced. Overall, he said, downtime can now be measured in minutes per year rather than hours.
Going forward the plan is to go live with a private cloud across all sites, using the VSPEX system to deliver the ERP system to the entire UK operations of MWW by the end of the year.
“What VSPEX is going to allow us to do is that we are going to be the data centre for all the other sites. We are actually going to pull it all into the server estate here, so we will end up hosting applications and running the entire UK operations IT from Birmingham."