.

The LAMP stack (Linux, Apache, MySQL and PHP / Python) is very popular among the new generation of Web 2.0 services. These services provide high availability and good performance for applications such as gaming portals, educational and corporate bookings e-commerce and travel, and many others. However, many issues such as performance database, hardware error bugs, application, denial of service attacks and other affect the performance of your service LAMP. The best approach is to key performance indicators to monitor and set rules for the notification, so you can react quickly to failures and attacks.

The first step of the installation is easy and without additional external monitors availability and performance, we recommend that external monitors Monit HTTP applications up Cover and dynamic pages and more important, the same for large files such as scripts and Flash objects. In addition, a simple ping facility the user can check whether the server is available. You also have the option of the router and firewall Ping, provides a way to determine where the fault lies in the whole system. SSH User can also set up monitoring SSH SSH if not, but the whole system remains active, you can isolate the problem specifically SSH. MySQL monitoring is an important part, and the user needs to ensure it is of the IP addresses Monit. DNA testing is a preventive measure against the problems of the beginning of your final DNS.

When the configuration of the external review of your application server, you need to establish the internal server resources. This monitoring will be on both troubleshooting and avoiding errors after installing the agent Monit your server (available Windows, Linux, Solaris, FreeBSD and Mac agents) are helping the following server monitors the Web interface Monit. Linux load average, CPU, disk and memory. The threshold for the average load should be set at 10-15 for the message. Set reasonable limits on memory and hard disk space and instructors. Apache / PHP and MySQL process should also be monitored. Set the threshold for CPU utilization per server per process (60-70%), especially for MySQL. It is often the cause for poor performance. Setting the low threshold can result in many warnings can cause too much alarm when the situation on the server is already critical.

For convenience, we recommend the use Monit application and service monitoring of transactions. It is very useful in multi-level complex applications such as online bookings, e-commerce, online banking and Web 2.0 services. It is related to external control, but is able to check the flow of the entire application in a control. Monit provides the Transaction Recorder Plug-in for your browser. In your browser you can save a script by navigating through the pages of the application and points beyond. Download the script on the server have Monit Monit check your application at regular intervals. In all controls combined Monit, 100% application availability for users is much closer to a reality.


Check availability