July 7, 2011, 11:18 p.m.
posted by alexpi
Keeping logs is a very important aspect of maintaining the security of your network, as logs can assist in everything from alerting you to an impending attack to debugging network problems. After an incident has occurred, good logs can help you track down how the attacker got in, fix the security hole, and figure out which machines were affected. In addition, logs can help with tracing the attack back to its source, so you can identify or take legal action against the intruder. In short, log files are worth their weight in gold (just pretend that bits and bytes weigh a lot). As such, they should be given at least as much protection as any other information that's stored on your servers—even the patent schematics for your perpetual motion machine.
This chapter deals mostly with various ways to set up remote logging, whether it be a simple central syslogd that your servers are logging to, setting up your Windows machines to send to a syslogd, or using syslog-ng to collect logs from remote sites through an encrypted TCP connection. Using these methods, you can ensure that your logs are sitting safely on a dedicated server that's running minimal services, to decrease the chance that the logs will be compromised.
Once you have all your logs collected in a central place, what can you do with them? This chapter also covers ways to summarize your logs into reports that are easy to read and understand, so you can quickly spot the most pertinent information. If that's not fast enough for you, you'll also learn how to set up real-time alerts that will notify you as soon as a critical event occurs. In some circumstances, responding immediately to an event—rather than waiting around for it to end up in a report that you read the next morning—can save hours of effort.