I am running the very latest updates of Proxmox 5.1 and the server does have KSM enabled, I had a VM kernel dump on me, and I noticed in the logs I am getting "page allocation failure" messages. Overall the RAM usage seems to be OK and the server is only running about 35 VMs each with 4GB of configured RAM. Please advise what you think I should do? Should I turn off KSM?
May 4 07:51:06 www kernel: kworker/0:1: page allocation failure: order:0, mode:0x310da
May 4 07:51:06 www kernel: CPU: 0 PID: 38 Comm: kworker/0:1 Not tainted 3.10.0-693.21.1.el7.x86_64 #1
May 4 07:51:06 www kernel: Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.11.0-0-g63451fca13-prebuilt.qemu-project.org 04/01/2014
May 4 07:51:06 www kernel: Workqueue: events_freezable update_balloon_size_func [virtio_balloon]
May 4 07:51:06 www kernel: Call Trace:
May 4 07:51:06 www kernel: [<ffffffff816ae7c8>] dump_stack+0x19/0x1b
May 4 07:51:06 www kernel: [<ffffffff8118cd10>] warn_alloc_failed+0x110/0x180
May 4 07:51:06 www kernel: [<ffffffff8118fa30>] ? drain_pages+0xb0/0xb0
May 4 07:51:06 www kernel: [<ffffffff816aa774>] __alloc_pages_slowpath+0x6b6/0x724
May 4 07:51:06 www kernel: [<ffffffff811912a5>] __alloc_pages_nodemask+0x405/0x420
May 4 07:51:06 www kernel: [<ffffffff811d5a38>] alloc_pages_current+0x98/0x110
May 4 07:51:06 www kernel: [<ffffffff8120241f>] balloon_page_enqueue+0x1f/0x90
May 4 07:51:06 www kernel: [<ffffffffc02b17b5>] update_balloon_size_func+0xc5/0x220 [virtio_balloon]
May 4 07:51:06 www kernel: [<ffffffff810abe2f>] process_one_work+0x17f/0x440
May 4 07:51:06 www kernel: [<ffffffff810acaf6>] worker_thread+0x126/0x3c0
May 4 07:51:06 www kernel: [<ffffffff810ac9d0>] ? manage_workers.isra.24+0x2a0/0x2a0
May 4 07:51:06 www kernel: [<ffffffff810b4031>] kthread+0xd1/0xe0
May 4 07:51:06 www kernel: [<ffffffff810b3f60>] ? insert_kthread_work+0x40/0x40
May 4 07:51:06 www kernel: [<ffffffff816c0577>] ret_from_fork+0x77/0xb0
May 4 07:51:06 www kernel: [<ffffffff810b3f60>] ? insert_kthread_work+0x40/0x40
May 4 07:51:06 www kernel: kworker/0:1: page allocation failure: order:0, mode:0x310da
May 4 07:51:06 www kernel: CPU: 0 PID: 38 Comm: kworker/0:1 Not tainted 3.10.0-693.21.1.el7.x86_64 #1
May 4 07:51:06 www kernel: Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.11.0-0-g63451fca13-prebuilt.qemu-project.org 04/01/2014
May 4 07:51:06 www kernel: Workqueue: events_freezable update_balloon_size_func [virtio_balloon]
May 4 07:51:06 www kernel: Call Trace:
May 4 07:51:06 www kernel: [<ffffffff816ae7c8>] dump_stack+0x19/0x1b
May 4 07:51:06 www kernel: [<ffffffff8118cd10>] warn_alloc_failed+0x110/0x180
May 4 07:51:06 www kernel: [<ffffffff8118fa30>] ? drain_pages+0xb0/0xb0
May 4 07:51:06 www kernel: [<ffffffff816aa774>] __alloc_pages_slowpath+0x6b6/0x724
May 4 07:51:06 www kernel: [<ffffffff811912a5>] __alloc_pages_nodemask+0x405/0x420
May 4 07:51:06 www kernel: [<ffffffff811d5a38>] alloc_pages_current+0x98/0x110
May 4 07:51:06 www kernel: [<ffffffff8120241f>] balloon_page_enqueue+0x1f/0x90
May 4 07:51:06 www kernel: [<ffffffffc02b17b5>] update_balloon_size_func+0xc5/0x220 [virtio_balloon]
May 4 07:51:06 www kernel: [<ffffffff810abe2f>] process_one_work+0x17f/0x440
May 4 07:51:06 www kernel: [<ffffffff810acaf6>] worker_thread+0x126/0x3c0
May 4 07:51:06 www kernel: [<ffffffff810ac9d0>] ? manage_workers.isra.24+0x2a0/0x2a0
May 4 07:51:06 www kernel: [<ffffffff810b4031>] kthread+0xd1/0xe0
May 4 07:51:06 www kernel: [<ffffffff810b3f60>] ? insert_kthread_work+0x40/0x40
May 4 07:51:06 www kernel: [<ffffffff816c0577>] ret_from_fork+0x77/0xb0
May 4 07:51:06 www kernel: [<ffffffff810b3f60>] ? insert_kthread_work+0x40/0x40
