swap_numa.rst (3100B)
1.. _swap_numa: 2 3=========================================== 4Automatically bind swap device to numa node 5=========================================== 6 7If the system has more than one swap device and swap device has the node 8information, we can make use of this information to decide which swap 9device to use in get_swap_pages() to get better performance. 10 11 12How to use this feature 13======================= 14 15Swap device has priority and that decides the order of it to be used. To make 16use of automatically binding, there is no need to manipulate priority settings 17for swap devices. e.g. on a 2 node machine, assume 2 swap devices swapA and 18swapB, with swapA attached to node 0 and swapB attached to node 1, are going 19to be swapped on. Simply swapping them on by doing:: 20 21 # swapon /dev/swapA 22 # swapon /dev/swapB 23 24Then node 0 will use the two swap devices in the order of swapA then swapB and 25node 1 will use the two swap devices in the order of swapB then swapA. Note 26that the order of them being swapped on doesn't matter. 27 28A more complex example on a 4 node machine. Assume 6 swap devices are going to 29be swapped on: swapA and swapB are attached to node 0, swapC is attached to 30node 1, swapD and swapE are attached to node 2 and swapF is attached to node3. 31The way to swap them on is the same as above:: 32 33 # swapon /dev/swapA 34 # swapon /dev/swapB 35 # swapon /dev/swapC 36 # swapon /dev/swapD 37 # swapon /dev/swapE 38 # swapon /dev/swapF 39 40Then node 0 will use them in the order of:: 41 42 swapA/swapB -> swapC -> swapD -> swapE -> swapF 43 44swapA and swapB will be used in a round robin mode before any other swap device. 45 46node 1 will use them in the order of:: 47 48 swapC -> swapA -> swapB -> swapD -> swapE -> swapF 49 50node 2 will use them in the order of:: 51 52 swapD/swapE -> swapA -> swapB -> swapC -> swapF 53 54Similaly, swapD and swapE will be used in a round robin mode before any 55other swap devices. 56 57node 3 will use them in the order of:: 58 59 swapF -> swapA -> swapB -> swapC -> swapD -> swapE 60 61 62Implementation details 63====================== 64 65The current code uses a priority based list, swap_avail_list, to decide 66which swap device to use and if multiple swap devices share the same 67priority, they are used round robin. This change here replaces the single 68global swap_avail_list with a per-numa-node list, i.e. for each numa node, 69it sees its own priority based list of available swap devices. Swap 70device's priority can be promoted on its matching node's swap_avail_list. 71 72The current swap device's priority is set as: user can set a >=0 value, 73or the system will pick one starting from -1 then downwards. The priority 74value in the swap_avail_list is the negated value of the swap device's 75due to plist being sorted from low to high. The new policy doesn't change 76the semantics for priority >=0 cases, the previous starting from -1 then 77downwards now becomes starting from -2 then downwards and -1 is reserved 78as the promoted value. So if multiple swap devices are attached to the same 79node, they will all be promoted to priority -1 on that node's plist and will 80be used round robin before any other swap devices.