Validating Client Cable Connections throughout the Data Center

Data center operations (DCO) require the right tools to test fiber-rich, carrier-neutral environments for redundancy, high-performance connectivity, and disaster recovery. As they scale their networks up to 400 Gbps, DCO and technicians struggle to install, verify, and troubleshoot client connections through passive and active data center interconnections.

Content Dam Lw Online Articles 2017 08 Lwanritsu080517fig4

Data center operations (DCO) require the right tools to test fiber-rich, carrier-neutral environments for redundancy, high-performance connectivity, and disaster recovery. As they scale their networks up to 400 Gbps, DCO and technicians struggle to install, verify, and troubleshoot client connections through passive and active data center interconnections. These connections can take many forms, such as direct attached cables (DACs), active optical cables (AOCs), single- and multimode fiber, and optical modules deployed almost daily in expanding data centers.

For these reasons, DCO require a hand-held, all-in-one tool for testing and troubleshooting electrical cables and fiber-optic interconnects between servers, switches, and routers at all data rates, using any given interface, located anywhere in the data center.

Interoperability Testing

As bandwidth demand grows, so do data center footprints, making it even more crucial to get higher-speed Ethernet technologies to the market as quickly as possible. Being able to test forward-looking Ethernet technologies in a vendor-neutral environment is key (see Table).

Lwanritsutable

Different Fiber-Optic Cables for Different Uses

DCO teams are responsible for the installation and verification of client cabling and transported services, from cloud applications to streaming content. Electrical Ethernet cabling such as active and passive DACs, Cat5e, Cat6, and Cat6a can be found in most plants for delivery up to 25 Gigabit Ethernet (GbE) rates for patch panels and horizontal or vertical cable managers.

Fiber-optic cables are common for most rates greater than 10GbE and can vary in types, connectors, and lengths depending on the application. MPO trunk cables can support up to 72 fiber strands, housed in one MPO connector, but are expensive in nature and used for short distances. Single mode fibers are used for longer distances (10-40 ,m) and higher data rates (up to 100 Gbps), but are cost-sensitive due to the optics required for data transmission. Multimode duplex fiber provides high data rates (up to 10 Gbps per fiber) over short to medium distances (up to 300 m).

For these reasons, data centers incorporate a combination of electrical and fiber-optic data connections, over both short and long distances, and use a variety of connectors depending on the type of network element, performance, and speed required to support the client service.

Fiber Optic Client Cable Testing

There are three main areas of fiber-optic client cable testing:

  1. Fiber-optic connector inspection
  2. Optical transceiver analysis
  3. Fiber-optic cable measurement

Let’s look at each in detail.

Connector inspection: How clean is the fiber connector? Dirty or damaged fiber-optic connectors can greatly affect network performance. Scratches, chips, and contamination on fiber-optic connector end faces reduce transmission quality and increase errors.

A digital video inspection probe can reduce these issues by verifying the condition and cleanliness of connector end faces during the installation phase without the need for analog microscopes. The connector image and detailed Pass/Fail status should be displayed as defined by IEC 61300-3-35, ensuring quality and consistency from connector to connector, technician to technician (Figure 1).

Lwanritsu080517fig1
Figure 1.The connector image and detailed Pass/Fail status should be displayed as defined by IEC 61300-3-35.

Optic transceiver analysis: Are my pluggable optical modules functioning correctly? As data rates increase to 25, 40, and 100 Gbps, pluggable optics and DACs become more complex and potential additional points of failure. Pluggable optical modules (in SFP, SFP+, SFP28, CFP, CFP2, CFP4, QSFP+, QSFP28, CXP, or other form factor) support management data input/output (MDIO ) and I2C information of the optical transceiver. I2C and MDIO analysis and testing verifies the performance of the pluggable optical moduel and monitors alarms and errors to isolate issues from the DAC and fiber-optic cables and connectors (Figure 2).

Lwanritsu080517fig2
Figure 2.MDIO analysis and testing verifies the performance of the pluggable optic and monitor alarms.

Fiber-optic cable measurement: What is my total loss? How much loss at each termination? Scratched or dirty connectors at fiber-cable connections, such as patch panels and adapters, can be detected as fault locations from the excessive optical reflections they create. An optical time-domain reflectometer (OTDR) displays such measurement results as a trace that shows the cable length, macro-bends, losses, and size of reflections, as well as an easy-to-view summary of the analysis results. The technician can set Pass/Fail conditions based on client service agreement value for each event, which are automatically identified within the OTDR trace. If the evaluation settings prescribed in the client agreement are set beforehand, the measured fiber-optic cable loss status can be easily distinguished visually when the measurement ends (Figure 3).

Lwanritsu080517fig3

Figure 3. An OTDR measurement displays results as a trace showing the fiber optic cable length.

Client Service Verification – Will the Data Connection Support the Client Service?

Selection of the proper benchmark verification test and location will determine client services through the data connection. Each benchmark test addresses a specific client service, demarcation point, or equipment to determine if the service-level agreement has been met (Figure 4).

Content Dam Lw Online Articles 2017 08 Lwanritsu080517fig4
Figure 4.Selection of the proper benchmark verification test and location will determine client services through the data connection.

Here are the most common benchmark tests:

  • RFC 2544 – This IETF standards body test typically is used for packet-based networkequipment or new lines. This traditional benchmark test is used to test the Layer 2 "pipe" throughput, latency, and frame loss after the physical Layer 1 fiber or electrical cable has been tested.
  • Y.1564 – This ITU standards body test is relatively new to the industry; it often is used to replace RFC 2544 in the operator end-to-end testing scenario but not for individual switch/router equipment. Y.1564 is a benchmark test used to emulate the "services" purchased by a client to verify the service acceptance criteria (SAC) agreed upon between client and service provider.
  • RFC 6349 – This IETF standard body test was recently developed. This benchmark test addresses the Layer 4 TCP communication within the packet-based transport. This test is focused on the client-to-client communication once the Physical, Transport, and Network layers are tested. Service providers and clients use this benchmark test to enable deeper testing into the packet; it is designed to test end-to-end throughput through a client demarcation point or firewall.

Conclusion

As DCOs test fiber-rich, carrier neutral environments and high-performance connectivity and disaster recovery, they require an all-in-one handheld tool for testing and troubleshooting electrical cables and fiber-optic interconnects between servers, switches, and routers at all data rates located anywhere in the data center. These tests can be achieved with the right handheld fiber-optic and client service verification tests.

Daniel Gonzalez is digital/optical business development manager for Anritsu Co. He possesses over 16 years’ experience in digital and optical transport testing, development, training, and execution spanning technologies including TDM, SONET, OTN, ATM, Carrier Ethernet, and Physical Layer signal integrity.

More in Network Test