Thanks!

We received numerous product and solution  awards recently. We also get a lot of recognition and media coverage. I would attribute those to the hard work, dedication and innovation  of our team. The latest nominee (which is actually an award) belongs to our customers.Net Optics has been selected by the editors of Inc. magazine as one of America’s fastest-growing private companies in the publication’s 2010 hall of fame.

Thank you!

Watch Net Optics CEO Bob Shaw on winning a spot on the 2010 Inc. 5000


Managing The Light Budget

One of the more important issues related to fiber optics based networking is proper management of the light budget, caused as a result of light loss inside fiber optics. A loss (power) Budget is the amount of attenuation that can be tolerated on the network and monitor links before the end-to-end data is corrupted.
Managing the light budget and understanding light loss is relevant for data centers as well as other FTTX (Fiber To The XX) networking. (See an interesting article here)

There are different reasons for light loss that occur during transmission of light signals inside fiber or during the interconnection process of two fiber-based components:

  • Absorption Loss: Particles or moisture in the fiber can block some of the light energy, it absorb the light and dissipate it in the form of heat energy, which caused absorption loss.
  • Rayleigh Scatter loss: Rayleigh scatter occurs at random when there are small changes in the refractive index of materials (the core and the cladding of the fiber optic cable) in which the light signal travels.
  • Bending Loss: occurs when a cable is bent and it disrupts the path of the light signal.
  • Return Loss is the loss of light signal that is reflected back to the original light source. This occurs as the light is reflected off the connector and travels back along the fiber to the light source.

The biggest loss often occurs due to insertion loss, which is the most important performance indicator of a fiber optic interconnection.  Insertion loss is the loss of light signal during the insertion of a fiber optic connector. It is measured in decibels (dB).  To maintain the minimum light budget loss, Net Optics Fiber Optics products (all different Tap and Tap modules) has different split ratio the amount of light that is re-directed from the network to the monitor ports) options.

To calculate and better understand split ratio, the Net Optics Fiber Split Ratio Reference Chart is available.

BTW, A good reference reading on light loss is provided by the Fiber Optic Association (see Guidelines On What Loss To Expect When Testing Fiber Optic Cables For Insertion Loss With A Meter and Source or OLTS).

– Sharon


Follow The Cable…

The other OMIf you were looking for another evidence of the dramatic shift in data center design and the migration to 10G, read the recently published “New Data Center Cabling Requirements” document that was released few days ago by Mark Fabbi(Gartner ID Number: G00205239).

This document, provides the physical cabling requirements for a growing (maybe I should write “big”) shift in network design.

In the past I spoke and wrote about the impact of 10G migration on the security pie market as it requires network and application security practitioners to look for new solutions that are capable to handle and/or secure and/or protect and/or monitor and/or inspect 10G (and beyond) traffic.

The design of the physical cabling infrastructure is affected by the migration from the current dominating school of end-of-row network architecture to the future top-of-rack switching architecture. Due to the use of 10G networks and future migration to 40G/100G, new data centers will require much less copper.

Fabbi recommends enterprises using a top-of-rack architecture to use copper in the rack and OM3 or OM4 fiber to connect to the core data center network. In addition, new data centers should overprovision multimode fiber cable by a factor of 4x in anticipation of a migration to 40 Gigabit backbones.

IMO, if you are involved with data center architecture and design, you should read this document.

Observations

I’m visiting Haifa, Israel this week. A place considered by many as one of the coolest places in the Mediterranean. (“cool” as in “hot”. I wish that it was colder).

This is a network and application security savvy country. Most of the customers that I am visiting are using more than a single tool to protect their applications. In some cases, there are more than two different tools to achieve higher level of security.

The tool manufacturers vary, but there is one thing in common: Tools are very often (I would say more than 70% of the cases) are all installed in in-line mode, blocking policy violations. A violation can be an attack, or a diverge from the corporate governance policy that might lead to information leak and data loss.  The corporations that I have visited are world leaders in their space. In other countries, there is more emphasis on business connectivity than in-line real-time, hard core blocking for security. In some cases, people refer to it as “monitoring” versus “preventing”.

Net Optics Director and Director Pro are used by those customers as a tool to ensure connectivity even when the highest degree of security is required. It lets them balance between multiple tools, to make each one more effective and route different types of traffic via an in-line route while allowing other types of traffic to be routes through a monitoring only device. Yet it also provides the means to very quickly (in a matter of less than seconds) to switch policies if needed.  One of the coolest use cases that customers are excited about is dynamic load balancing for inline applications using Director Pro. This unique solution allows to run multiple security device in in-line mode for blocking while avoiding asymmetric routing issues.


University of Bradford Workshop: The Need to Monitor Speed

Greetings from Bradford.

Tomorrow I’ll be presenting at a workshop hosted jointly by the University of Bradford and Net Optics partner, MarQuest Ltd., This is a one-day event that focuses on high-speed networking for security and performance. Workshops will cover such issues as minimizing analysis tool costs, performance-related security in high-speed networks, and developing an effective monitoring infrastructure.

The event features Professor Demetres Kouvatsos Head of the Networks and Performance Engineering Research Group (NetPEn), Informatics Research Institute (IRI), University of Bradford. Chris Bloom, WildPackets Principal Software Engineer/ Developer Evangelist, Chris Devereux, Motorola Senior Technical Architect and myself.

The topics covered will concentrate on the challenges of monitoring and analyzing data from networks which transport gigabit data streams. This will include the following areas:-

  • Developing an effective monitoring infrastructure.
  • Minimizing analysis tool costs.
  • Techniques for network forensics.
  • Performance Related Security in High Speed Networks.
  • Monitoring security in high speed wireless infrastructure.

I have developed a new presentation trying to keep it “vendor free” while presenting my thoughts on developing the right infrastructure for effective monitoring. One of the thoughts that I will be presenting is related  to the different tradeoff between the different functions derived by business decisions:

  • Performance vs. Accuracy
  • Security vs. Connectivity (my favorite)
  • False Positive vs. False Negative
  • Price vs. Features.
  • Ease of use vs. Availability and features

I’ll post the slides after tomorrow.


Reminder: T1, T3

A European friend is now performing is first  implementation in the US and as such is starting to “discover” the differences between T1 and E1. Being such a nice person (what are friends for), here’s a short reminder:

T1

  • A T1 line is a pure digital line.
  • The T1 line can handle 24 simultaneous voice calls.
  • A T1 line can handle a high speed Internet connection speed of 1.544 Mbps.
  • Or, as an integrated service it can handle a combination of fewer voice calls with some bandwidth given to Internet connections.
T3
  • A T3 line is also a pure digital line. It is the equivalent of 28 T1 lines or 672 POTS lines.
  • A T3 line can handle 672 simultaneous voice calls or provide a high speed Internet connection of 44.736 Mbps.

See also the network dictionary for a comparison between T1 and E1. Other useful information here.


Best Practices For 10G Migration

Tomorrow (well, technically today) we will be hosting the first event in a serias of technology seminars in our offices in Santa Clara. Tomorrow’s topic is 10G migration. We are expecting a large number of attendees and looking forward for a great, interactive discussion.

Hope to see you there!

8:30-9:00 – Registration and Continental Breakfast
9:00-9:15 – Greeting and Welcome
9:15-10:15 –Best Practices for 10G Migration: Sharon Bessor – VP of Technology
10:15-10:30 – Coffee Break
10:30-11:15 – Guest Speaker: Sean Power – Co-author of Complete Web Monitoring
11:15-12:00 – Use Cases & Demonstration: Dennis Carpio – Director of Product Innovation 
12:00-1:00 – Catered Lunch
1:00-2:45 – Group/Panel Discussion
2:45-3:00 – Break and Net Optics Manufacturing Tour
3:00-3:30 – Final Thoughts & Wrap-up

Thoughts The Enterprise LAN Magic Quadrant

Last week (6.10) I read Gartner’s Magic Quadrant for Enterprise LAN (Global) that was written by Mark Fabbi and Tim Zimmerman. As always, Gartner provides interesting insights that reflect what the large customers are doing. Since Net Optics is operating in an adjacent market, I have the luxury of providing my thoughts without the associated risk.

  • Market size: I’ll take Gartner word for it, but I am very interested to learn how they measure the market and came with the revenue number for 2009. In my opinion the number that is quoted in the MQ is lower than it should be.
  • I totally agree with the observation regarding the increase in demand for 10G, though in my opinion it is now being used for more than interconnections between switches.
  • I wish that there were more emphasis in vendor evaluations on 40G and 100G Ethernet vision.
  • Low latency, Top-of-Rack (TOR) and emerging new Ethernet standards are now a fact of life. The data center IS changing. Blame it on Virtualization, cloud or cost of electricity and air conditioning, it does not matter…

In addition, Gartner made some interesting comments regarding the changes in the landscape as a result of economic conditions (changes in the buy patterns and habits of large organizations ) as well as the M&A activity (HP acquired 3Com and Avaya acquired all of Nortel’s enterprise networking business, including its LAN switching portfolio) or lack of  (Brocade).

If you are working for a large organization or plan to refresh your enterprise LAN or build a new data center, you should read this paper.


New White Paper: Gaining Total Visibility for Lawful Interception

The term “lawful interception” (LI) describes the process by which law enforcement agencies conduct electronic surveillance of circuit and packet-mode communications as authorized by judicial or administrative order.

Each country has its own criteria for defining criminal or terrorist activities; the purpose of LI is to gather evidence to detect and prevent such activities. In the United States, the Communications Assistance for Law Enforcement Act (CALEA) sets forth how service providers must support lawful interception. Abroad, the European Telecommunications Standards Institute (ETSI) drives adoption of standards for telecommunications, broadband, and related technologies in Europe and other countries.

Read this whitepaper here. Other resources are available here.


It’s Happening and It’s Makes Sense

I just read an interesting Enterprise Strategy Group (ESG) white paper (see PDF here) that was commissioned by Emulex. This document makes a lot of sense and is echoing a growing industry trend that we are witnessing: Data Center consolidation.

I will let you read this document, which reflects a real phenomenon observed in the field, though in my opinion puts too much emphasis on FC consolidation.

Based on a previous ESG survey (2010 IT Spending), top investments will focus on increased use of server virtualization, security, data center consolidation, upgrading networks, and protecting and managing data.
Organizations are investing in network in order to ensure access. Networks provide the necessary infrastructure that without it, an organization would not function. In today’s economy, organizations are trying to improve their advantages while reducing costs thus consolidation that provides both cost benefits as well as competitive advantages makes sense (duh…).

One area of disagreement with this white paper content is around the focus of consolidation and where should a converged network be deployed first. The white paper suggests that initial deployments include blade centers and unified computing environments ad well as top of the rack switches (ToR) for new environments.  While I agree that it makes sense (duh again…) I think that ToR deployments makes more sense to begin with.