The network has been designed to accommodate clients demanding the highest quality network performance.

 

There is a central focus on redundancy allowing the network to rapidly self-heal failures without interruptions to connectivity.

 

 

The redundancy is multi-tiered with N+1 internal device elements as well as entirely redundant chassis allowing any routing device to fail without interrupting client data connectivity. All core routing and switching equipment is state of the art Cisco.

Tier-1 Premium Bandwidth Providers

Providing the highest quality Internet connectivity and performance is our top priority. Fitting with this premium Tier-1 bandwidth providers have been exclusively selected, ensuring minimal latency and excellent reach ability on a global basis. We are proud to feature direct connectivity with MCI/UUnet, AT&T, and Savvis/Cable & Wireless. The aggregate connectivity reaches an astounding 4,500+ networks just a single network away from our uplinks.

SBC Diverse Path OC-48 Fiber

Fiber connectivity to the facility is supplied by an SBC OC-48 diverse path fiber optic sonet ring. Featuring level 7 path diversity the sonet ring has multiple entry points into the facility as well as connectivity directly to multiple carrier offices. There is no shared physical path at any point in the entire sonet loop. This design allows for uninterrupted service even in the event of a fiber cut, or other direct CO equipment failure. The optical equipment is capable of supporting multiple OC-192 connections.

Cisco GSR 12000 Border Routers

Network connectivity begins at our border layer which connects directly to the Tier 1 service providers as well as the network core. A reliance on multiple devices each with internally redundant systems ensures the highest levels of reliability.

Each Cisco GSR is capable of independently routing 60 gigabits of data at wire rate while isolating any fault risk to each independent line card. Connections to the core are derived independently enabling yet another layer of redundancy and performance load balancing.

Cisco 6509 Sup720 Core Network

The fully redundant network core handles bandwidth aggregation and routing for the entire network. Each Cisco 6509 Sup720 is capable of routing up to 400 million packets per second with a total capacity of 720 gigabits per second each. This layer is fully redundant with each unit possessing the ability to route the entire core in the event of a device hardware or software failure. Core failure recovery is a nearly instantaneous process and is fully automated by the self-healing core elements.

Cisco 3550 Distribution Switches

Final layer3 distribution to rack switches as well as VLAN segmentation is handled by Cisco Catalyst 3550 distribution switches. Each switch is capable of handling simultaneous wire rate data traffic for every port configured ensuring optimal performance and eliminating blocking potential to the core or individual rack switch.

At the rack layer we utilize Hewlett-Packard Procurve switches to follow and further enforce VLAN segmentation protocols from the distribution layer as well as offer final customer network access.

Datacenter Power Systems Overview

The power systems feature extensive fault tolerance and resilience at every layer. Incoming service is routed underground to a dedicated on-site transformer. This system routes to a transfer switch junction which also connects to the power generation facilities. The primary UPS is a 300kva Powerware 9315 featuring redundant battery cabinets and a full maintenance bypass allowing for service and upgrades without interruption of power to our servers. Liebert Precision power distribution units handle final power transformation and distribution to racks.

Generator backup services are provided to our facility by a Cummins 313kVa Diesel Generator featuring over 24 hours of autonomous runtime before requiring refueling. Generator power is activated automatically in the event of a grid failure by the transfer switch. The intermediate load is capable of being handled for over an hour by the UPS, however this is not necessary as the generator is active and up to speed within 16 seconds of a power failure.

Facility General Environment

The facility includes over 11,000 square feet of available space. Construction is built to surpass local building area codes and seismic requirements. Industry standard hard floors are covered with vinyl tile and overhead wiring systems are utilized for enhanced cooling and a more secure cable infrastructure.

Datacenter access is strictly limited to technical staff. Electronic security systems control datacenter access and are accompanied by a full compliment of motion detecting security cameras which are strategically placed throughout the entire facility. Furthermore the facility external walls are concrete poured to a height of twenty feet.

Liebert 22 Ton AC Unit

Environmental processing systems include redundant Liebert Precision 22 ton up flow air conditioning units. Temperature and humidity are precisely regulated year round to ensure optimal equipment reliability. Each unit contains independent compressors and cooling loops to further enhance fault tolerance and reliability.

Air filtration systems actively remove foreign particulates from circulation and cycle the entire datacenter air supply in a matter of minutes. This is a fully managed facility, as such technical staff includes level 3 technicians on site 24 hours per day, allowing incident response times to be kept to a minimum and system integrity retained with utmost urgency.

Safeguard Computer Services, Inc. utilizes the services of Liquid Web, Inc. to manage the data center and hardware maintenance.