10 troubleshooting tips

Network managers share the essentials any data center shouldn't be without.

Everyone who backpacks can rattle off the 10 essentials by rote - map, compass, flashlight, extra food and clothes, sunglasses, first-aid kit, knife, waterproof matches, fire-starter. They are the critical and essential tools anyone who hikes needs as insurance against the unexpected.

Likewise, network managers rely on must-have troubleshooting tools for diagnosing problems with the network, servers and desktop PCs. What follows are the 10 tools network managers deem essential to their network management arsenals.

1. Configuration inventory. An inventory of the configuration of workstations and servers on the network is instrumental in being able to root out the source of trouble, according to Mark Moroses, senior director of technical services and security officer at Maimonides Medical Center in New York. "This can be on paper, but it cannot become suspect - the data has to be maintained and trusted," he says. "It is priceless to know the exact configuration of equipment inside the problem domain."

To continue reading, register here and become an Insider. You'll get free access to premium content from CIO, Computerworld, CSO, InfoWorld, and Network World. See more Insider content or sign in.

Everyone who backpacks can rattle off the 10 essentials by rote - map, compass, flashlight, extra food and clothes, sunglasses, first-aid kit, knife, waterproof matches, fire-starter. They are the critical and essential tools anyone who hikes needs as insurance against the unexpected.

Likewise, network managers rely on must-have troubleshooting tools for diagnosing problems with the network, servers and desktop PCs. What follows are the 10 tools network managers deem essential to their network management arsenals.

1. Configuration inventory. An inventory of the configuration of workstations and servers on the network is instrumental in being able to root out the source of trouble, according to Mark Moroses, senior director of technical services and security officer at Maimonides Medical Center in New York. "This can be on paper, but it cannot become suspect - the data has to be maintained and trusted," he says. "It is priceless to know the exact configuration of equipment inside the problem domain."

Moroses uses Bindview's NETinventory for this purpose. Another popular option is Tally Systems' TS Census, a product that Novell recently acquired and added to its ZENworks Asset Management.

2. Network diagramming software. "You cannot troubleshoot a network issue without an accurate, complete understanding of the equipment involved, and for that you must be sure of the communication path," Moroses says.

Network diagramming software is key to establishing the critical links between network gear. Microsoft's Visio 2003 is the market leader for diagramming the physical configuration of the network. Users can download a trial version here.

3. A device polling system: A simple piece of software that automatically polls network devices - servers, switches, printers and routers - to see if they are connected also is important. When a device doesn't respond to a poll, it is usually the first signal of trouble.

"Polling is an accurate, clear method for assessing the scope of the problem," Moroses says. "For example, multiple devices in multiple locations may be affected at the same time."

Moroses says that relying on users to complain that a printer isn't working or that they can't connect to the network is usually misleading and results in IT staff taking longer to diagnose the problem. While SNMP traps can be useful, simple ping polling is worth its weight in gold, Moroses says.

Ronald Godine, manager of IS operations for Royal Appliance in Glenwillow, Ohio, recommends SolarWind's Network Management Toolset. "A copy of SolarWinds will let you interact with a variety of devices and map ports on a switch," he says.

Many devices such as Cisco routers and switches also ship with a device polling configuration option. Fidelia's NetVigil also provides for device polling of any SNMP-enabled device.

4. Application logs: Each application such as Oracle provides a log file of error codes. At Maimonides, Moroses says that he has found that by reviewing the logs early in the troubleshooting process he can reduce the size of the fault domain. He says application logs are especially effective when used in combination with network device logs. Most applications generate logs of errors that will tell the IT manager that a process terminated and for what reason.