Company: TimesSquare² Case Study
Overview:TimesSquare² is an alliance between Thompson-Reuters and NASDAQ that sells outdoor media on a massive LED display system in Times Square NYC. The signs on both the NASDAQ building and the Thompson-Reuters building can play synchronized video or operate independently.
The Reuters sign is the largest digital display system in the world, spanning 22 stories and 7,000 square feet at 3 Times Square. The Nasdaq sign, known as the Nasdaq MarketSite Tower, is seven stories tall and contains 10,000 square feet of signage.
The system that runs the signs is highly customized consisting of eight application servers, four database servers, six administrative servers and twenty video display machines. All of the hardware is near the end of its life expectancy and the annual cost to maintain warranties has increased significantly. The original design specifications did not take expansion into consideration.
Requirements:TimesSquare² requested that a multi-phased approach be taken to completely overhaul the system. In the first phase, the application needed to be modified to accommodate multiple sign displays. Connectivity between the application network and the physical sign with NASDAQ needed to be changed. The monitoring and alerting systems also needed to be modified to notify the new partners for appropriate events associated with the NASDAQ sign.
The goal of phase two was to replace the existing hardware with new equipment. In doing so, TimesSquare² asked for network and application design changes that improved performance, availability, and lowered operating costs. The final recommendation also needed to provide a means to display an alternate video source on any sign should the primary source experience a failure.
- Install connectivity between the NASDAQ sign data center and the core data center where the application is housed
- Create a method to display video on the NASDAQ sign should the connectivity between data centers fail
- Use a combination of a SAN and virtual server technology to provide fault tolerant services and access to data.
- Upgrade and modify existing monitoring software to predict and detect failures. Use this information to take appropriate preventative and corrective actions while alerting the appropriate parties.
- Install a video matrix switch that will allow dynamic redirection of video signals that can be triggered automatically or via the monitoring software.
The DetailsPhase One: We installed point-to-point fiber between the data center for the NASDAQ sign and the data center for the application that provides network and video connectivity. Additional fiber strands are installed for redundancy.
The application was modified to allow additional signs to be managed should the need arise. Sign display machines were installed at the NASDAQ sign data center that displays generic video templates in the event connectivity is lost to the main data center.
Phase Two: By using dual display video cards and a slight modification to the application, the number of video display machines required was cut in half. Each display machine can effectively display content on 2 separate sign pieces.
We were able to replace sixteen of the servers by migrating over to virtual machines. Three physical virtual servers now simulate the applications and services originally provided by sixteen servers. In addition, a majority of the storage required has been moved to a fiber channel SAN which provides high availability and fault tolerance. The SAN, when coupled with NPIV technology, provides great flexibility in implementing clustered services and data on the virtual machines.
A video matrix switch was installed that controls all video sources and outputs. The switch is programmed with preset configurations that allow for common situations including maintenance and failures. The video matrix has RS-232 connections that allow signals to be sent to it from either the custom application or the monitoring software.
Square One installed a newer version of the management software to monitor the availability and performance of the hardware as well as the processes on the equipment. Metrics have been defined and used to create action plans for recovery. In addition, this software is used to automate and schedule various maintenance processes on both the servers and video display machines.