Crafting an automation strategy that gives back
A comprehensive collection of articles, videos and more, hand-picked by our editors
When a host operating system is short on physical memory, the hypervisor polls balloon drivers installed on the guest operating systems to request available memory. The guests respond by using page reclaiming algorithms to determine which pages are available and can be assigned to the balloon drivers. The process of assigning available pages to the driver is known as inflating the balloon. Releasing available pages is known as deflating the balloon.
In a positive light, memory ballooning permits the total amount of RAM required by guest virtual machines to exceed the amount of physically available RAM on the host. A problem can occur, however, when a balloon driver inflates to the point where the guest VM no longer has the memory it needs to run processes within the guest. In such a scenario, the guest operating system will start swapping things out to deal with its memory issues. This swap activity can negatively affect performance, depending upon the amount of memory to recoup and/or the quality of the storage IOPS delivered to the VM.
Continue Reading About memory ballooning
'memory ballooning ' is part of the:
View All Definitions