Guest

Quality of Service (QoS)

Cisco QoS Technology Plays Major Role in NASA Network

Customer Case Study

Text Box: EXECUTIVE SUMMARYCUSTOMER NAME● NASA Jet Propulsion Laboratory, Deep Space Network●   Pasadena, CaliforniaBUSINESS CHALLENGE●   For more efficient and cost-effective use of bandwidth, NASA's Deep Space Network (DSN) managers wanted the ability to provide bandwidth guarantees to a large number of different traffic streams. NETWORK SOLUTION ●    Using Class-Based Weighted Fair Queuing (CBWFQ), Multilink Point-to-Point Protocol (MLPPP), Low Latency Queuing (LLQ), and Link Fragmentation and Interleaving (LFI) in Cisco IOS Software, DSN managers have been able to create 16 different traffic classes.●  These Cisco technologies have provided NASA with minimum bandwidth guarantees and the ability to enforce strict priority for voice over IP (VoIP) traffic, resulting in predictable network jitter.BUSINESS RESULTS●  By introducing CBWFQ, MLPPP, LLQ, and LFI in NASA's DSN WAN, administrators were able to gain precise control over how their expensive international WAN resources are utilized.

NASA uses quality of service technology in Cisco IOS® Software to precisely manage its international WAN bandwidth.

Monitoring 35 concurrent missions in low-Earth orbit and deep space, the United States National Aeronautics and Space Administration (NASA) Deep Space Network (DSN) carries a steady flow of data between rovers, landers, and orbiters, and their respective operators using tracking complexes around the world. To cost-effectively utilize this network, which costs an estimated US$10 million annually to operate, NASA network engineers have defined 16 different classes of service that are given differing network bandwidth guarantees using Class-Based Weighted Fair Queuing (CBWFQ) in Cisco® IOS® Software version 12.2(27). The agency's goal was to achieve better control of bandwidth utilization and to assign quality of service (QoS) commitments for voice-over-IP (VoIP) traffic.

BUSINESS CHALLENGE

NASA's DSN is headquartered at the Jet Propulsion Laboratory (JPL) in Pasadena, California. JPL is a federally funded research and development center, administered by the California Institute of Technology. For more than 40 years, the network has supported the communications of space missions ranging from Earth orbiters to Mars Rover explorations and probes of Jupiter and Saturn. More than 25 years after it was launched, the Voyager spacecraft continues to send data, which is now encapsulated into Internet Protocol (IP) on the DSN ground network.
Giant satellite dishes measuring up to 70 meters in diameter are located at three Deep Space Communications Complexes (DSCCs)-in the United States, Australia, and Spain. The WAN infrastructure at each facility includes twin Cisco 3640 modular access routers connected to three diversely-routed T1 dedicated circuits connected to the Central Communications Terminal (CCT) located at JPL. Network administrators use CiscoWorks to manage the network and use virtual local area networks (VLANs) to provide location-independent subnetworks for mission-support teams. The DSN ground network is a network based largely on Cisco routers and switches.
The DSN supports dozens of individual flight projects. DSN ground network managers needed a way to manage their WAN bandwidth in a way that was consistent with the commitments that they were making to their network users. The network management team is responsible for a LAN on the campus in Pasadena and for all communications between devices located at the remote DSCCs and devices and mission support teams located at the CCT in Pasadena.

NETWORK SOLUTION

Previously, priority queuing (PQ), was being used, but there were limitations in how traffic could be recognized. There were only four queues to use, and the higher priority queues had the potential to completely starve the lower priority queues. Using CBWFQ, the DSN network managers created 14 data traffic classes, plus default and voice, for a total of 16. They can now ensure that each traffic class receives at least its minimum bandwidth requirement. The traffic policies were defined using bandwidth percentages, and applied to the Multi-Link Point to Point Protocol (MLPPP) interface, so if a bundled circuit is lost, all the traffic classes are affected proportionately.
CBWFQ extends the standard weighted fair queuing algorithm to allow the creation of up to 64 user-defined traffic classes. With CBWFQ, a share of a link can be allocated to each class of traffic using either relative bandwidth percentages (which is useful in environments where interface speeds can change dynamically) or fixed bandwidth values given in kilobits per second. Traffic can be classified by the type of service byte (DiffServ field in the DiffServ model) by checking the IP precedence bits or differentiated service code point (DSCP) bits in the IP packet header, or by a simple or extended access control list check. It is even possible to use logical combinations of the above two techniques. The bandwidth assigned to a class is the guaranteed bandwidth that will be delivered during periods of congestion. Characteristics such as weight and maximum packet limit may also be assigned to a class.
During situations where there is unclaimed bandwidth, the remaining bandwidth is divided up according to the relative weights of the minimum bandwidth guarantees. The DSN network managers keep track of the traffic classes using a Microsoft Excel workbook that shows the CBWFQ percentages and kbps calculations. The network managers can tell their users what their bandwidth guarantees would be in scenarios where one or two of the three T1s in the MLPPP bundle fail.

BUSINESS RESULTS

NASA proved the usefulness of CBWFQ and captured it in Figure 1, showing a 20-minute traffic blip graphed on a four-hour trending chart. A spacecraft entering the range of an Earth ground station began transmitting high-priority flight project data (red line) that has a 2 Mbps maximum transmission rate. This flow displaced the lower-priority Very Long Baseline Interferometry (VLBI) data based on measuring quasar noise (cyan line).

Figure 1. CBWFQ Balances High and Low Priority Traffic


 

Traffic Class

Last

Average

Maximum

 

Total Traffic

216.15 kb/s

3.31 Mb/s

3.70 Mb/s

 

VLBI

0.00 b/s

2.88 Mb/s

3.38 Mb/s

 

Flight Project Data

115.53 kb/s

281.88 kb/s

2.12 Mb/s


Once the mission-critical flight project data playback stopped, the VLBI stream resumed its prior bandwidth utilization level, which was limited to a maximum of 3 Mbps using traffic policing. In addition to CBWFQ, the DSN uses low latency queuing (LLQ) and link fragmentation and interleaving (LFI) to guarantee voice quality. LLQ provides the best of both queuing worlds: the fairness of weighted fair queuing and the strict priority queue from priority queuing, which is used exclusively for VoIP traffic.
The DSN network managers are planning to use Cisco's Network Based Application Recognition (NBAR) technology to differentiate between multiple traffic streams utilizing the same Transmission Control Protocol ports and IP addresses. This helps enable more granular traffic classification so that they can manage how network resources are being used by each flight project.
Managing bandwidth with this level of precision has contributed to savings in the DSN's international WAN circuit expenditures by allowing network managers to optimize current bandwidth utilization as opposed to requesting funding for bandwidth increases.

FOR MORE INFORMATION

To find out more about Cisco Solutions and Services, visit: http://www.cisco.com.
To find out more about quality of service in Cisco IOS Software, visit: http://www.cisco.com/go/qos.
This customer story is based on information provided by the United States National Aeronautics and Space Administration (NASA) Deep Space Network and describes how that particular organization benefits from the deployment of Cisco products. Many factors may have contributed to the results and benefits described; Cisco does not guarantee comparable results elsewhere.
Cisco provides this publication as is without warranty of any kind, either express or implied, including the implied warranties of merchantability or fitness for a particular purpose. Some jurisdictions do not allow disclaimer of express or implied warranties, therefore this disclaimer may not apply to you.