Dynamically optimizing memory allocation across virtual machines

A method for allocating memory among a plurality of computing systems, said method comprising:

• assigning a free memory threshold to each computing system wherein at least one free memory threshold assigned to a computing system is determined so that one of a probability of the computing system achieving paging does not exceed a threshold and a probability of the plurality of computing systems achieving paging, does not exceed a threshold;
• periodically monitoring each computing system for memory usage;
• and
• in response to said monitoring determining that an amount of free memory for the computing system is below the free memory threshold for the computing system, allocating memory from a free memory pool to the computing system.

A method for allocating memory among a plurality of computing systems, said method comprising:

• periodically monitoring each computing system for memory usage;
• in response to said monitoring determining that a computing system is running low on free memory, determining if a memory size of the computing system has recently increased;
• and
• in response to one of determining that the memory size of the computing system has not recently increased and determining that the memory size of the computing system has increased at a rate falling below a threshold, allocating memory from a free memory pool to the computing system.

A method for allocating memory among a plurality of computing systems, said method comprising:

• assigning a free memory pool size threshold to a free memory pool, wherein memory from the free memory pool can be allocated to a plurality of the computing systems;
• dynamically varying the free memory pool size threshold in response to requests to the free memory pool for memory;
• periodically monitoring each computing system and the free memory pool, for memory usage;
• in response to determining that a size of the free memory pool is below the free memory pool size threshold, identifying at least one computing system with sufficient free memory;
• and
• transferring memory from said at least one computing system to the free memory pool.

A method for allocating memory among a plurality of computing systems, said method comprising:

• assigning a free memory pool size threshold to a free memory pool, wherein said free memory pool size threshold is determined so that a probability of the free memory pool running out of memory does not exceed a threshold;
• periodically monitoring each computing system and the free memory pool for memory usage;
• in response to determining that a size of the free memory pool is below the free memory pool size threshold, identifying at least one computing system with sufficient free memory;
• and
• transferring memory from said at least one computing system to the free memory pool.

A method for allocating memory among a plurality of computing systems, said method comprising:

• periodically monitoring each computing system and a free memory pool for memory usage, wherein memory from the free memory pool can be allocated to a plurality of the computing systems;
• in response to said monitoring determining that the free memory pool is running low on free memory, determining if a size of the free memory pool has recently increased;
• in response to one of determining that the size of the free memory pool has not recently increased and determining that the size of the free memory pool has increased at a rate falling below a threshold, identifying at least one computing system with sufficient free memory;
• and
• transferring memory from said at least one computing system to the free memory pool.