Managing the extended enterprise: remote monitoring tools can improve productivity and make best use of limited resources.
Because more than 45% of network problems occur at the physical layer, the ability to proactively monitor and manage physical Ethernet connections over extended distances can help IT managers keep their networks running and their users productive. Those managers need tools for predicting, troubleshooting and fixing Layer 1 problems remotely. Instead, slow and costly onsite diagnostics are often used, perhaps necessitating a second onsite visit to correct the problem and fully restore the network.
The two most common ways to determine the health of remote equipment are IP ping and SNMP. The ping provides a simple method for contacting a remote device and determining whether that equipment is able to respond, with the downside that little information is returned.
SNMP provides far richer data and, with vendor extensions, can provide environmental as well as operational data for managed devices at the remote end. Typically, SNMP operates inband to discover problems in the network and their root causes, and then send critical management information (as SNMP traffic) onto a link that is already suspect.
The network slows as it continues to search for the problem, and herein lies the "catch 22" in today's remote-management solutions. IT managers need to identify problems without adding more burden to the network or depending on the already compromised physical network to transport the management data required to fix that same link.
STANDARDS ON THE WAY
So, although SNMP offers many benefits in terms of remote management, it incurs long polling delays, which increase in direct proportion to the size of the network. To be effective, remote SNMP management solutions need the ability to monitor links in real time without utilizing user bandwidth, creating latency or impacting network efficiency.
The IEEE is currently working on a standard, 802.3ah Ethernet in the first mile, which limits the amount of user bandwidth used during polling, alleviating network degradation, which can occur during an SNMP poll The challenge, however, is to have the remote device "push" information to the management station when certain thresholds are reached, minimizing the impact on network performance. This would allow the network manager to identify problems and schedule maintenance before a catastrophic network failure occurs.
What is needed is a solution that enables efficient management of the physical layer and allows network technicians to see and test a remote link without utilizing user bandwidth, monitor equipment at a remote site for temperature power and voltage, measure the optical power transmit and receive, and analyze data traffic for loss of packets and errors-without a visit to the remote site.
TOOLS FOR TROUBLESHOOTING
IT managers need the ability to reach across the entire enterprise to communicate, test and reconfigure an unmanaged remote device, without reducing available bandwidth to the individual user sites. To accomplish this, they need sophisticated management capabilities, including remote loopback testing, real-time quality-of-line (QoL) statistics, quality of equipment monitoring (QoE), quality of optical amplitude monitoring, and dynamic bandwidth provisioning. All of this information needs to be available in a standards-conformant management format that is transmitted without impacting user bandwidth or network performance, and stored in a database that can be used to determine trends and for predictive analysis.
Remote loopback testing is a critical troubleshooting tool that can reduce operational costs by eliminating costly and time-consuming onsite visits by scarce IT technicians.
Up-to-date QoL data must he collected and displayed in real-time to enable IT personnel to analyze user usage, predict needs based on information kept in a trends database and make dynamic adjustments to accommodate traffic on an ongoing basis, thus ensuring optimal network throughput and performance.
QoE is necessary to ensure high availability and can be used for remote monitoring of power levels and equipment temperature, so that potential problems can be detected and corrected proactively. The IEEE 802.3ah "dying gasp" specification allows determination of where a problem has occurred. Dying gasp checks the remote site for a power problem before a remote site complaint occurs.
Should the signal start to degrade without a dying gasp or remote site complaint, 802.3ah will initiate a "discovery restart" to indicate that there are errors on the link. While this feature does not isolate the problem, it does alert the network manager that problems are starting to occur.
Quality of optical amplitude monitoring mitigates the problems associated with optical link troubleshooting by supporting remote measurement of the optical power on fiber links and remote monitoring of" the quality of optical amplitude. Any optical power deviations that occur after maintenance or new installations may indicate that fiber connectors were not installed properly. While this may not signal an immediate problem, the administrator can add these checks to the scheduled maintenance.
Because today's IT organizations have smaller staffs to manage and troubleshoot each connection, uptime becomes uncertain, employees become frustrated and companies lose tens of thousands of dollars in productivity. Yet, with the right tools, nearly all of these problems can be addressed.
For more information from Metrobility: www.rsleads.com/407cn-264
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Remote Management|
|Date:||Jul 1, 2004|
|Previous Article:||SIP routing.|
|Next Article:||Consider remote-control options.|