I just want to comment this issue I was having that was not easy to come across due to the amount of lines on /var/log/messages, and maybe helps someone else.
I deployed my cluster using this repo, and my k8s workers starting freeze completely and not respond to any action, the only solution was to restart them but when begins to start the pods it freeze again, so I needed to remove them before launch them again.
That issue was happening on Oracle Linux 7.4 using Docker 17.06.2.ol and k8s 1.8.14
The solution was just increasing this value sudo sysctl fs.inotify.max_user_watches=1048576
I just want to comment this issue I was having that was not easy to come across due to the amount of lines on
/var/log/messages, and maybe helps someone else.I deployed my cluster using this repo, and my k8s workers starting freeze completely and not respond to any action, the only solution was to restart them but when begins to start the pods it freeze again, so I needed to remove them before launch them again.
That issue was happening on
Oracle Linux 7.4usingDocker 17.06.2.olandk8s 1.8.14The solution was just increasing this value
sudo sysctl fs.inotify.max_user_watches=1048576