IT Infrastructure Management for Data Centers and Network Optimization

Written by
john.jensen

With technology evolving it can be a challenge for CSP's to optimize existing facilities and networks in time to meet demands. Optimization can mean a variety of projects when it comes to IT infrastructure. Managing these projects while new technologies such as vCCAP, FTTH, 5G, and Remote PHY are being deployed is a challenge the industry is facing. Optimization of existing networks is required because these new networks need to be planned and designed a long side our existing networks.

There is risk of site expansions when platform specific equipment is not strategically placed and therefore, can require optimization projects. Some examples of innovative network planning to avoid these risks are: using a modular approach, use modern compact equipment, plan the equipment in a way in which you minimize cabling and losses, and proactive equipment maintenance planning. Other opportunities in data center and headend optimization projects include adjusting cooling set-points of the facility, lighting and airflow management, power de-centralization, and predictive design:

 

Adjusting Cooling Set-Points

There is a potential cost savings in energy consumption by slightly adjusting the cooling set-points of the CRAC units in a given data center or headend facility. If planned correctly with air-flow management, applying a few CRAC unit settings can impact energy consumption and reduce operating costs.

Airflow Management

The new technologies keep increasing the power density per rack. This requires proactive planning in containment to separate cool from warm air for proper airflow management.

No alt text provided for this image

 

Power De-Centralization

Over-crowding power cable trays is an issue that arises with increasing energy demand to accommodate these new technologies. BDFB’s take up a lot of space in headend facilities. Power de-centralization is required to take advantage of significant savings by removing the big bulky legacy BDFB’s and installing small (sometimes rack-mountable) DC Plants that can feed mini BDFB’s or even connect directly to rack mounted fuse panels.

Predictive Design

Having a ‘predictive analytics’ approach to IT infrastructure management can create a 50% savings in design costs. Traditional reactive based projects are what cause a lot of the issues the industry faces with their current networks evolution. A modular approach is definitely the preferred route to take with any new builds to truly see the advantages of predictive analytics in IT infrastructure management.

No alt text provided for this image

A mesh of new technologies will be active at the same time in the near future as we approach 5G and evolution of cable networks. Remote-PHY is being used right now as the ‘in-between’ technology to migrate existing networks customers to the new technology, with higher bandwidth and speed requirements. 5G and FTTH will force the traditional wireline and wireless networks to adapt. It directly affects the way we look at IT infrastructure management and how we plan for the future in our existing networks. 

No alt text provided for this image

 

Therefore, we really need to focus on having an optimization mindset during not only the design phase of new projects or new facilities, but also during the installation of new technologies in existing networks. The space, power, and cooling trifecta requires a new future proof planning approach with predictive analytics to prepare for the IT load demand. 

 

John Jensen, Network Technical Lead 
Missing Link Technologies

Find John's Article on LinkedIn

 

RELATED CONTENT:

 

Datacenter, Optimization, airflow management, predictive design

Comments

*

Restricted HTML

  • Allowed HTML tags: <a href hreflang> <em> <strong> <cite> <blockquote cite> <code> <ul type> <ol start type> <li> <dl> <dt> <dd> <h2 id> <h3 id> <h4 id> <h5 id> <h6 id>
  • Lines and paragraphs break automatically.
  • Web page addresses and email addresses turn into links automatically.