One of the largest multinationals operating in the building materials (notably cement production) space with presence in many countries of the world and a large scale cement manufacturing factory in Northern Nigeria.
-The client has a broad range of applications hosted in the global HQ datacenter that needed to be accessed as quickly as possible with the barest minimum downtime as inputs from quantities of materials about to be dispatched must be sent before being approved, important emails concerning running the plant and oiling the business processes must be read and attended to promptly and a host of other application services which need to be accessed by users on the Lan. Also voice calls on the VoIP network had very poor quality of service and problems arisng that could have been sorted out over the phone necessitated physical movements and in turn lost time of manpower. It was very easy to see how holes in the IT system were negatively impacting the business processes and in turn profit margins.
Southgate Technologies Limited was brought in based on our stellar track record of turning around comatose IT systems around for the better. Fact-finding trips were carried out to access the issues and a team of reputable engineers including in-house Cisco CCIEs worked on the design of the solution. The challenges amidst others were identified as:
-multiple cascaded layer 2 networks which increased the broadcast domain and decreased network performance.
-quality of service and congestion issues which affected voice over ip networks.
-low speed switching based on low speed interfaces connecting to critical resources.
-general network best practices that needed to be adjusted.
To fix all these we approached the challenge from 2 angles:
- network devices: 3 Sets of devices were introduced into the network: 1) Cisco Catalyst 4506 Switches 2) Cisco Catalyst 3850 Switches and 3) Cisco 2960 Catalyst Switches working on the Core, Distribution and Access layers of the network respectively.
- network architecture: the network was redesigned from a flat cascaded layer 2 network into a 3 tiered network (core, distribution and access layer) where routing was confined to the core while switching were confined to distribution and access, this cut down the broadcast domains tremendously while switches also pruned vlans not in their path, spanning-tree protocols was seriously tweaked to get the best convergence upstream and to have instant access downstream to hosts, the fastest possible speeds were used to switch at interfaces and in some cases interface bundling was carried out for faster speed, QoS queuing and congestion avoidance policies were implemented and security best practices were enforced.
- Faster access to emails which sped up business making decisions.
- Crisp and clear voice quality on the VoIP network.
- Access to applications hosted in the global HQ was faster, 0% packets dropped over an extended monitoring period and latency far lower than before, all of this meant cement could be dispatched far faster than before based on faster and more reliable inputs hence increasing profits.
The largest electricity producer in a central African country with presence in all states of the country.
The client has an on-premise datacenter with hosted Netapp storage and obsolete servers and network devices all operating in silos looking to transition their datacenter to support 21st century needs, host hundreds of applications and collapse all the silos into a single datacenter infrastructure.
Southgate Technologies Limited was contacted to help the company in this complex transition to a more agile IT process where the siloes of storage, compute and network are merged into one singular solution. Netapp’s Flexpod Converged Infrastructure (comprising Cisco Network, Netapp Storage, Cisco Compute and Microsoft Virtualization technology) had been purchased and Southgate was brought in on a professional services basis for implementation.
Our implementation approach involved: 1) breaking down the converged infrastructure into its different components (compute, storage, network, virtualization and management) to design and implement 2) coupling all the different component together 3) testing the assembled infrastructure based on acceptable parameters developed with the client.
The network section consisted of the Cisco Nexus 5548UP switch, compute section consisted of Cisco UCS B200 M4 blade servers, Virtualization consisted of Microsoft’s Hyper-V, management of Cisco UCS Director and Netapp storage.
- Lesser heat and power consumption since all the individual siloes now work together as a single unit not separately and on a more of
- Less cabling due to cisco’s unified architecture that combines both Lan and San traffic on same cable instead of the usual separation.
- Faster access to applications since all the interfaces was 10Gb/s switching architecture which was a significant upgrade from the prior 1Gb/s and Nexus’ NX-OS superior QoS implementation and lossless connectivity.
- Multi-tenancy: the datacenter was built with the ability to segment different sections for traffic isolation and segmentation and so if issues arrive from applications deployed in the test and development environment it doesn’t affect the core traffic.
- All of these ultimately lead to faster and more agile business processes and satisfaction for clients and higher profit margins.