Isolating and resolving network problems Posted on February 9, 2012 by Tim Burke When it comes to keeping your IT network performing optimally, finding and fixing network faults certainly is crucial. But it’s just as crucial to sustain network operations while the issue gets resolved. Which means that the first order of business is to ensure that your network has entered one of the failure modes you’ve designed for it so that remaining network resources are allocated according to your organization’s business priorities. Two approaches are used to resolve network problems: Event reconstruction and status analysis. Which network administrators use depends on whether the network’s problem is ‘soft’ or ‘hard’. Resolving network problems when automation isn’t enough Problem type Example Resolution approach Requirements for resolution approach Hassle quotient Soft Network congestion (abnormal line/node condition, excessive load due to unexpected traffic or security breach) Event reconstruction: Reviewing alert messages and analyzing events as they occur Log of network events with highly accurate timestamps on all alert messages Time-consuming Hard (problem state persists even when network is in failure mode) Node fault (power failure, equipment failure) Trunk failure (fiber cut) Network status analysis Remote diagnosis as much as possible Sufficient facilities for testing (e.g., line problems require testing device endpoints, device problems require testing each interface) Manual repair onsite is costly as well as time-consuming and lengthens time of network abnormality Whatever the nature of your network problem, be sure to log it carefully and thoroughly so network administrators’ response can be analyzed and their procedures tuned. That will minimize cost and downtime should the problem recur. Read more about this on Quest’s networking monitoring solutions page.