|Network Diversity and Performance
Our network has been engineered from the ground up to accommodate the high-availability demands of the mission-critical systems we manage. Our Juniper-powered, Dual Routing Engine Equipment has unique self-healing attributes that allow us to deliver a near 100% infrastructure availability guarantee. The resulting architecture is redundant in every element, from edge routers to core switches. In the event that one of these components should fail, the redundant network device will pick up the functionality without interruption of service. The convergence time is around 50ms and is practically hitless. Our network architecture is designed with dual physical routes to our providers, as well as the physical hardware itself which is connected to our backbone and peering / transit providers. The equipment provides high availability features for maximum network uptime by providing carrier class reliability and redundancy. Our network is fully meshed and redundant with multiple backbone providers using Border Gateway Protocol (BGP) from multiple diverse routes: Internet Solutions, MTN, Neotel, SAIX, Telia Sonera and Savvis to mention a few.
Gigabit VLAN Customer Network Layer
Our customer network layer utilizes multiple redundant Juniper switches with Virtual Chassis technology combining the high availability and carrier-class reliability of modular systems with the economics and flexibility of stackable platforms, which enable us to deliver a high-performance, scalable solution for our data centre customers. Flexible GbE and 10GbE uplinks to our Juniper MX core enable high-speed connectivity for aggregation on our collocation environment. The top of the range solutions with in our collocation environment provides automatic router backup and 100% uptime SLA, by using Virtual Router Redundancy Protocol (VRRP), without having to use a dynamic routing or routing discovery protocols, for customer gateways. Each customer, depending on their requirements, can be provisioned a separate VLAN providing traffic isolation and security.
Customers have an option of being provided with a shared Firewall facility to port filter and protect against intrusion detection within our collocation environments, which is also a fully redundant piece of equipment on our network. Alternatively customers can be provided with a managed Juniper firewall or provide and manage their own firewall for protection.
VLANs are used to increase the security of Vox Telecom customer network environments and provide segregation and isolate customers to their own networks. Our Juniper EX switches utilizes Layer 2 (L2) VLANs which have the ability to provide additional security in a network environment. Client’s network traffic is isolated within their VLAN allowing no other client to see other client’s network traffic.
Points of Presence
Currently through our initial network design, Vox Telecom has a scalable design that allows us to radically facilitate our growth and allows us to have predictable costs over time. Currently our Points of Presence are:
- London Telehouse
- London Telecity
- Teraco Isando
- Rosebank (Jinx)
- Pretoria / Centurion
- Teraco Rondebosch
- Bree Street (Cinx)
- Westville (DBN)
- Teraco Riverhorse Valley
- Port Elizabeth
- East London
- Richards Bay
Vox Telecom also has Points of Presence in CINX (Cape Town Internet Exchange) and JINX (Johannesburg Internet Exchange). The Vox group has invested heavily in infrastructure and networking facilities to generate economies of scale. Maximising the number of minutes carried reduces cost, and the savings achieved are passed directly on to Vox’s customers. The new international cables coming into South Africa will drop bandwidth prices as well as making immense amounts of new capacity available, and Vox is in the perfect position to deliver that profitably to its customers.
Please see below a schematic of the Vox Network with all the peering links around the country:
Neotel offers next generation voice, data and Internet services. Additional services include Virtual Private Networks, hosting and satellite services. Their various telecommunications licences, Electronic Communication Network Service License and Electronic Communication Network License, allow them to provide the entire range of telecommunications services.
Broadlink (Pty) Ltd is a subsidiary of Wireless Business Solutions Holdings offering a last mile telecoms alternative through the use of licensed microwave technology.
Telkom is Africa's largest integrated communications company, and provides integrated communications solutions
Metro Ethernet and Logical Wireless
Logical Wireless provides integrated satellite communication solutions to enterprises across all industries.
The MTN Group Limited is a leading provider of communication services, offering cellular network access and business solutions.
Vodacom delivers end-to-end converged solutions, extending from mobile to fixed line voice and data, managed data networks, Voice over IP (VoIP) solutions, hosted facilities and applications, security and managed hosting solutions.
iBurst offers a wireless alternative to fixed line connectivity.
Quality of Service
-to-end, right from a customer’s VoIP gateway, through the access and core networks, all the way to the termination gateway or device. In our multiservice environment, we use cross-service QoS classification alignment and consolidation with planning policies to ensure that the backbone network is able to ensure priority traffic at all times.
To effectively and predictably deliver the differing Quality-of-Service (QoS) requirements of voice, video, and data applications, we leverage the application-level QoS and traffic management features and mechanisms of our Cisco and Juniper routers to control bandwidth, delay, jitter and packet loss. We categorize our traffic into multiple classes (Premium, Gold, Silver, Bronze), with QoS parameters defined for each class.
The Vox Datapro Environment currently offers the following QOS/COS Levels:
- AF (Data) (Contention 1:1, Latency = <40ms)
- BE (Data) (Contention 2:1, Latency = <80ms)
- LBE (Data) (Contention 3:1, Latency = <150ms)
Vox Datapro implements Platinum Service (AF) for all our Backbone MPLS Customers to ensure fastest response between sites.
Our Service Centre is a separate operating division of Vox Telecom Limited and addresses the Vox Telecom Group’s needs in terms of customer servicing as well as network monitoring and escalation procedures.
The Service Centre operates on a 24/7/365 basis and facilitates incoming calls on a “branded” basis should this be required, customers are all identified separately and dealt with within their brand services. In addition we facilitate the option of a separate incoming contact number for logging, prioritization and escalation allowing for more specialised services. The Service Centre’s technical desks are based on skill levels to ensure timeous and efficient resolution of customers’ service requests (SR) as well as ensuring that the correct skill sets are managing the SR’s.
Monitoring and Event Management
We have an extensive suite of operational support software which collects and mines CDRs to report on service quality and alert threshold exceptions to our Operations and Maintenance staff as necessary. Devices also alert via SNMP trap to centralized systems management platforms when events occur, such as MOS score levels dropping below acceptable values, interfaces going down, interface errors, etc.
Our MPLS Environment provides detailed monitoring and management, measurable in an SLA (Service Level Agreement). Full Network statistics are available via MRTG Graphing. Customers have access to a full Invoice history as PDF downloads and can change passwords and manage contact details via our Customer Zone Interface.
All critical systems are operated in 1+1 hot standby mode, with no or minimal downtime experienced when the primary unit fails. Hardware and Security redundancy solutions are additional to the proposed solution, and only implemented at the customer’s request.
There are several methods that can be used to achieve a ‘zero service loss’ upgrade. The method depends mainly on the transmission media used to serve the site. With certain transmission types, the provider will interrupt the service during the upgrade. In these cases one would need to install a temporary alternate transmission path, perhaps via microwave radio. But if the connectivity is via Ethernet over fibre, the upgrade can be hitless.
Our preventative maintenance policy is designed to minimize customer impact. Where preventative maintenance action requires downtime, we notify our customers with at least 7 days notice. Such downtime is then scheduled to occur when service usage is at its lowest demand level, which is normally between 2-4 am on Sunday mornings. If the maintenance action impacts only a single customer, we negotiate a mutually acceptable maintenance window. Such service impacting preventative maintenance actions is a very uncommon occurrence. Our critical systems are designed such that maintenance can be performed without impacting customer service.
Hardware Refresh Cycles
Our hardware refresh cycles are typically on the order of 5-10 years, depending on the pace at which technology developments make it beneficial to do so. If cost savings can be achieved for our customers by upgrading our products, or replacing aging systems, we do so. If customers require services enabled by innovations available through new products, we integrate these into our network. We also track the equipment MTBFs and supplier support and use this information in our decision making processes w.r.t. technology refresh.