Using the Anturis Console, you can set up monitoring of free physical memory for any hardware component (a server computer) in your infrastructure by adding the Free Physical Memory monitor to the component.

Create Your Account >

Physical memory (also known as main memory or primary storage) is the only type of memory directly accessible by the CPU. It is used to store data that is actively being processed, as well as instructions that the CPU reads and executes. Data can be stored and retrieved in any order, which is why physical memory is often referred to as random-access memory (RAM), although auxiliary memory (such as, optical discs, magnetic disk drives, flash memory and other secondary storage devices) are also accessed in a random-access manner.

It is also very important for the CPU to access data in the physical memory in the shortest time possible, regardless of its location on the medium. RAM used for primary storage is fast, but volatile, meaning that data is only stored while there is power, and cleared at startup.

Running out of free physical memory is one of the reasons for server performance degradation. Some systems use secondary storage as virtual memory, moving the least-used data from the physical memory and retrieving it back when it is required. Besides the fact that secondary storage is much slower, such swapping leads to file system fragmentation, which contributes to an even greater decrease in server performance.

When there is little RAM left, you may want to consider optimizing the way physical memory is used by the OS and other software. If you are not able to reduce the amount of used physical memory, then you should add more RAM to the server.

A memory leak is a common problem for server software. It usually happens due to poor design, when an application does not properly discard unused objects from the main memory. The amount of memory constantly increases until there is no memory left for new objects, and the application crashes. RAM monitor can help identify a memory leak early and properly react to this problem before a crash occurs.