Kubespray: Meet a install failure issue, error message: "assertion": "ansible_swaptotal_mb == 0"

[wumingpu]

Is this a BUG REPORT or FEATURE REQUEST? (choose one):

Environment:

• Cloud provider or hardware configuration:
  My localhost VM

• OS (printf "$(uname -srm)\n$(cat /etc/os-release)\n"):
  Linux 4.4.0-62-generic x86_64
  NAME="Ubuntu"
  VERSION="16.04.2 LTS (Xenial Xerus)"
  ID=ubuntu
  ID_LIKE=debian
  PRETTY_NAME="Ubuntu 16.04.2 LTS"
  VERSION_ID="16.04"
  HOME_URL="http://www.ubuntu.com/"
  SUPPORT_URL="http://help.ubuntu.com/"
  BUG_REPORT_URL="http://bugs.launchpad.net/ubuntu/"
  VERSION_CODENAME=xenial
  UBUNTU_CODENAME=xenial

• Version of Ansible (ansible --version):
  ansible 2.4.2.0
  config file = /root/.kubespray/ansible.cfg
  configured module search path = [u'/root/.kubespray/library']
  ansible python module location = /usr/lib/python2.7/dist-packages/ansible
  executable location = /usr/bin/ansible
  python version = 2.7.12 (default, Nov 20 2017, 18:23:56) [GCC 5.4.0 20160609]

Kubespray version (commit) (git rev-parse --short HEAD):
c2347db

Network plugin used:
flannel

Copy of your inventory file:

[kube-master]
master

[all]
master ansible_host=10.22.1.180 ansible_user=root ip=10.22.1.180
node1 ansible_host=10.22.1.181 ansible_user=root ip=10.22.1.181
node2 ansible_host=10.22.1.182 ansible_user=root ip=10.22.1.182
node3 ansible_host=10.22.1.183 ansible_user=root ip=10.22.1.183

[k8s-cluster:children]
kube-node
kube-master

[kube-node]
node1
node2
node3

[etcd]
master

Command used to invoke ansible:
time kubespray deploy --verbose -u root -k /root/.ssh/id_rsa -n flannel

Output of ansible run:

task path: /root/.kubespray/roles/kubernetes/preinstall/tasks/verify-settings.yml:76
Wednesday 06 December 2017 16:51:43 +0800 (0:00:00.161) 0:00:24.714 **
fatal: [master]: FAILED! => {
"assertion": "ansible_swaptotal_mb == 0",
"changed": false,
"evaluated_to": false
}
fatal: [node1]: FAILED! => {
"assertion": "ansible_swaptotal_mb == 0",
"changed": false,
"evaluated_to": false
}
fatal: [node2]: FAILED! => {
"assertion": "ansible_swaptotal_mb == 0",
"changed": false,
"evaluated_to": false
}
fatal: [node3]: FAILED! => {
"assertion": "ansible_swaptotal_mb == 0",
"changed": false,
"evaluated_to": false
}
to retry, use: --limit @/root/.kubespray/cluster.retry

---

< PLAY RECAP >

---

    \   ^__^
     \  (oo)\_______
        (__)\       )\/\
            ||----w |
            ||     ||

localhost : ok=2 changed=0 unreachable=0 failed=0
master : ok=17 changed=0 unreachable=0 failed=1
node1 : ok=15 changed=0 unreachable=0 failed=1
node2 : ok=15 changed=0 unreachable=0 failed=1
node3 : ok=15 changed=0 unreachable=0 failed=1

Anything else do we need to know:

[jicki]

cd /tmp

rm -rf node*

在每台服务器上执行

swapoff -a

执行完毕以后~ 再重新运行 程序

[ipeacocks]

I could confirm this issue. Swap on VM is turned off

# free
              total        used        free      shared  buff/cache   available
Mem:        1796368      544636      110716        8484     1141016     1070604
Swap:

but Ansible installer of kubespray doen't think so

TASK [kubernetes/preinstall : Stop if swap enabled] *************************************************************************************************************
Sunday 10 December 2017  03:26:41 +0200 (0:00:00.057)       0:00:12.176 ******* 
fatal: [k8s-s1.me]: FAILED! => {
    "assertion": "ansible_swaptotal_mb == 0",
    "changed": false,
    "evaluated_to": false
}
fatal: [k8s-m1.me]: FAILED! => {
    "assertion": "ansible_swaptotal_mb == 0",
    "changed": false,
    "evaluated_to": false
}
fatal: [k8s-m2.me]: FAILED! => {
    "assertion": "ansible_swaptotal_mb == 0",
    "changed": false,
    "evaluated_to": false
}

Ubuntu 16.04, baremetal installation.

[ipeacocks]

Fully recreated VMs - and issue has gone away. I suppose it was due to cached Ansible facts or smth like that.

[jicki]

@ipeacocks rm -rf /tmp/node*

[ipeacocks]

@jicki, yes, I've tried that but it didn't help me.

[ipeacocks]

I suppose this is correct path for cleaning cached facts:

/home/<your_username>/.ansible

Or even better recreate facts each run but it needs changes to code.

[wumingpu]

@jicki 谢谢，问题解决了，我先在Master上rm -rf /tmp/node，然后每台VM上执行swapoff -a。再运行程序就没那个错误了，谢谢。
@ipeacocks The issue resolved, I run "rm -rf /tmp/node" on Master and then run "swapoff -a" on each VM. Then rerun the deploy command, I do not get this issue. You can try to run "swapoff -a" command on each VM.

[hobbytp]

yes, I met the same issue, my VM is centos, after remove /tmp/node* in Ansible server, and disable swap (remove swap in /etc/fstab and swapoff -a and reboot), the issue gone. Thanks for the info sharing.

[junaid-ali]

Removing /tmp/node* and disabling swap worked for me. @hobbytp we don't need to reboot since swapoff -a will disable swap immediately. For persistence, we do need to update /etc/fstab

[junaid-ali]

I am a beginner in writing playbooks but there should be a cleaner way here instead of reading facts from /tmp directory.

[Ghostbaby]

/tmp/node* 不在每个节点上，在kubespray所在服务器上。