Alert Monitoring system: if the data center of an enterprise obtains the most advanced set of new monitoring tools, it has all functions and even performs related technologies. When a staff member can look at each system and process for your organization, it feels a sense of accomplishment.
Buying an expensive tennis racket does not mean to be ready for the Wimbledon open, which is the same for monitoring. The right tools do not make any sense to those who don't use it. And knowing when and how to create meaningful, operational and valuable alerts for businesses and businesses can make the data center monitoring of enterprises more effective.
test
Before you start creating new data center monitors and alerts, correct tests should be performed. Here are some tips to keep in mind:
At the time of testing, reduce the scope as much as possible and expand slowly. What the enterprise really wants to test first is the trigger condition when the alarm is issued or the alarm is triggered.
Learn to use the reverse threshold. Enterprises need to avoid duplication of testing machines. CPU<90% will trigger more reliably.
Record and log again. If the enterprise tools support their own log records, please open them. In the enterprise alarm operation, insert as many updates as possible. From "I start developing step X" to "I have just completed the step X", it may seem too long, but when we try to find out where the process is in the 17 step process, we will be happy to do that.
Experience with yourself. Make sure you receive alerts and don't try to uninstall them until the number and quality of the message is the message that the enterprise is willing to accept.
Avoid email. Sending alerts via email only increases the latency and infrastructure pressure. Instead, messages should be sent to local log files to display and so on.
Collaboration. The staff will have to endure the results of new monitors or alerts, so that everything from basic functions to message formats should be agreed.
Introduction
Now that you know how to test it, it's time to talk about what to test, and more specifically, where to start.
It makes sense to focus on the environment with the least effort to provide the maximum reward. This may mean listening to other teams complaining about the IT problems they face, which may help to determine whether the problems they encounter are driven by system failures. Then, complex alerts can be set up.
Although it sounds paradoxic, it doesn't have to go too far. Once more teams start to find help, they may find themselves in an awkward position.
profit
It is important to remember to monitor the actual purpose of improvement before people begin to use the new monitoring tools. The goal of IT professionals is to save business capital.
One of the challenges of good data center monitoring is to develop, test and persuade time spent by business managers, which is valuable and profitable.
So, how to show the digital statistical proof of the success of monitoring? One option is to review the past, that is, to keep the ticket data before the new monitoring and alarm is in place. By showing how high the number of tickets before the new system is, and how low it is, how much time can be saved for new surveillance and alerts. The color of the computer is better than anyone who knows that time is money.
According to people's experience, there are three drivers to help managers make decisions: increase income, reduce costs, and avoid risks. If the proposed project, the proposal or the software does not select any of these boxes, you can now give up.
Fortunately, effective monitoring can meet at least two of these requirements. It reduces cost by allowing IT professionals to find and solve problems faster, thus avoiding expensive downtime and reducing business expenses.
Monitoring also helps to avoid risk. An integrated monitoring tool that provides in-depth data will help IT professionals actively respond to potential failures. Effective monitoring is not waiting for the occurrence of outage events and after the remedies, but effective monitoring, so that people can avoid the problem.
Monitoring is not only the right assistant of the IT professionals, but also the key factor for enterprises to seek to reduce the risk and cost. However, this is not a quick solution. In order to get the rich rewards of monitoring, people must first learn to use it correctly.
沒有留言:
張貼留言