最近我遇到一个问题，我在带有 k8s 的 docker 上运行 mysql，系统负载在 2 天内从 1->500+ 升高，/var/log/messages 信息：
Aug 17 19:50:07 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:07.435900074+08:00" level=debug msg="Calling GET /v1.26/containers/json?all=1&filters=%7B%22label%22%3A%7B%22io.kubernetes.docker.type%3Dpodsandbox%22%3Atrue%7D%7D&limit=0"
Aug 17 19:50:07 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:07.442853242+08:00" level=debug msg="Calling GET /v1.26/containers/json?all=1&filters=%7B%22label%22%3A%7B%22io.kubernetes.docker.type%3Dcontainer%22%3Atrue%7D%7D&limit=0"
Aug 17 19:50:07 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:07.703904603+08:00" level=debug msg="Calling GET /v1.23/containers/json?limit=0"
Aug 17 19:50:07 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:07.804893547+08:00" level=debug msg="Calling GET /v1.26/containers/json?filters=%7B%22label%22%3A%7B%22io.kubernetes.docker.type%3Dpodsandbox%22%3Atrue%7D%7D&limit=0"
Aug 17 19:50:07 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:07.809574497+08:00" level=debug msg="Calling GET /v1.26/containers/json?all=1&filters=%7B%22label%22%3A%7B%22io.kubernetes.docker.type%3Dcontainer%22%3Atrue%7D%2C%22status%22%3A%7B%22running%22%3Atrue%7D%7D&limit=0"
Aug 17 19:50:08 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:08.427098775+08:00" level=debug msg="Calling GET /version"
Aug 17 19:50:08 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:08.427932286+08:00" level=debug msg="Calling GET /version"
Aug 17 19:50:08 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:08.428708265+08:00" level=debug msg="Calling GET /v1.26/version"
Aug 17 19:50:08 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:08.453153388+08:00" level=debug msg="Calling GET /v1.26/containers/json?all=1&filters=%7B%22label%22%3A%7B%22io.kubernetes.docker.type%3Dpodsandbox%22%3Atrue%7D%7D&limit=0"
Aug 17 19:50:08 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:08.456364735+08:00" level=debug msg="Calling GET /v1.26/containers/json?all=1&filters=%7B%22label%22%3A%7B%22io.kubernetes.docker.type%3Dcontainer%22%3Atrue%7D%7D&limit=0"
Aug 17 19:50:08 SVR5798HW1288 systemd[1]: Starting Cleanup of Temporary Directories...
Aug 17 19:50:08 SVR5798HW1288 dockerd[1890]: time="2018-08-17T19:50:08.656418404+08:00" level=debug msg="Calling GET /v1.26/version"
Aug 17 19:50:08 SVR5798HW1288 systemd[1]: Started Cleanup of Temporary Directories.
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] Linux version 4.10.13-1.el7.elrepo.x86_64 (mockbuild@Build64R7) (gcc version 4.8.5 20150623 (Red Hat 4.8.5-11) (GCC) ) #1 SMP Thu Apr 27 12:06:06 EDT 2017
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] Command line: BOOT_IMAGE=/vmlinuz-4.10.13-1.el7.elrepo.x86_64 root=/dev/mapper/VolGroup00-lv_root ro crashkernel=auto net.ifnames=0 biosdevname=0
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point registers'
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] x86/fpu: Supporting XSAVE feature 0x002: 'SSE registers'
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] x86/fpu: Supporting XSAVE feature 0x004: 'AVX registers'
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] x86/fpu: xstate_offset[2]: 576, xstate_sizes[2]: 256
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] x86/fpu: Enabled xstate features 0x7, context size is 832 bytes, using 'standard' format.
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] e820: BIOS-provided physical RAM map:
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] BIOS-e820: [mem 0x0000000000000000-0x000000000009d3ff] usable
Aug 19 08:30:51 SVR5798HW1288 kernel: [ 0.000000] BIOS-e820: [mem 0x000000000009d400-0x000000000009ffff] reserved
8 月 18 日完全没有任何问题，我于 8 月 19 日 08:30 重新启动系统，因为系统冻结，没有内核恐慌并且没有日志，我的问题是，systemd-tmpfiles-clean.service 是否可能导致系统挂起？我已经遇到过两次这个问题，不知道该怎么办。
内核：4.10.13
systemd 版本：219