Navy to test how network outages affect NMCI

The Navy is testing a service that can easily show how network outages, both planned and unplanned, affect mission readiness.

The Pacific Fleet (PACFLT) command commissioned EDS to develop the service, called the Common Operational Picture (COP), for the Navy Marine Corps Intranet (NMCI), which EDS manages.

The idea behind the service is to identify the specific mission areas and even individuals who are affected when a network service is disrupted for some reason.

PACFLT commissioned the service last fall, and EDS was able to put a production version in place by December for the PACFLT command center. EDS plans to develop other user-defined operational pictures that could be used elsewhere in NMCI, which would allow the integrator to offer the service to the rest of the Navy and the Marine Corps commands.

Usually when network performance falters, information technology personnel characterize the problem as the first step of fixing the problem. And when routine maintenance is performed, all the components being serviced are well documented.

But this information only flows within the IT department. Non-IT personnel who use these systems may not be aware of the outages. Moreover, commanders may not fully understand how the downtime affects their ability to carry out operations.

"We have people who stand watch over the network. But they could not really give that real-time status when their higher authorities required it because they were unable to understand what [what impact a given incident had] on the network," said Navy Cmdr. Everett Hayes.

The COP is designed to bridge the worlds of IT notifications and the rest of the command. Each time an outage occurs, it generates a list of personnel and missions affected by the issue.

EDS developed a system that would allow the Navy to define operational pictures for missions and personnel. The integrator then linked these pictures back to the equipment used on the network, so when the equipment performance would be impacted by some event, an alert would be sent as to how the incident would affect the personnel and missions.

The nice thing about these alerts is that they describe adverse network events "in a language" that all personnel can understand, Hayes said.

Core services that are being monitored by include checking e-mail and being able to access remote applications. Issues that might affect these services could be a server that has stopped working or an upgrade of some software or firmware that requires a rebooting of its host machine.

At times, even planned outages can have an effect on users in non-obvious ways. "In a network like NMCI, planned maintenance is huge undertaking because there are so many different components that are affected," said Mark White, who is the EDS NMCI account manager for enterprise services. An upgrade of a Microsoft Exchange server in San Diego may affect personnel in Hawaii who use that server for e-mail.

"Let's say you have a significant position at the PACFLT command and wanted to know whenever an outage or an unplanned outage would affect you — to keep you from being able to use the network in the full capacity, or limit some of the things that you might need to complete your task or job," Hayes said. "In the user-defined COP, we furnished EDS with your name, and they mapped that to certain devices in then network. If one of those devices was affected by a planned or unplanned outage, it would come up on the COP and say that you would no longer be able to send e-mail, or you were being affected by latency issues, [or whatever the problem was]."

Such knowledge "would give the people on watch a kind of agility and make [them] more responsive," to mission needs, Hayes said. Elsewhere in the Navy, outages are conveyed to end-users from an NMCI change-management team, but typically they lack the detailed insight into operations to tell how specific personnel and missions would be impacted.

With the COP, personnel can view incidents that may impact them through a Web browser or by an e-mail alert, White said. The service also details recent changes in network configurations. The COP keeps track of a description of the event, the status of the efforts to remedy or complete the event, the time of the event, the reason for the event, and its the operational impact. When NMCI IT personnel log a network-disrupting event, the COP system automatically generates a list of which users are affected.

To build the service, EDS tied together a number of commercial network-monitoring applications, White said. They include Netscout's Network Intelligence Suite, which includes the nGenius InfiniStream for network-traffic analysis, and the Cisco Network Analysis Module software for the company's Catalyst switches.