Does CRI-O support live upgrade without restarting containers?

cn flag

I have an on-prem kubeadm created Kubernetes cluster running on Ubuntu 18.04. I have installed cri-o as container runtime. Since, it is claimed by CRI-O that it supports zero downtime during live upgrades, as it doesn't restart the containers when the CRI-O service is restarted.

In my cluster, when I restart cri-o service (systemctl restart crio), all the containers are restarted. Is there any specific configuration that I am missing that will avoid the restart of the container when the cri-o daemon is restarted?

SYN avatar
hk flag
You are correct, CRI-O containers should not restart with service. Maybe check your service script, any chance it would be the one shutting down your containers? Can you share CRI-O version, configuration, maybe the steps you've followed setting it up? I doubt it would help, but have you tried with a recent OS/kernel -- 18.04 is getting old

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.