Upgrading containers from 3.4 to 7.1, possible?

[PicklesTheOtt]

I have an old, OLD cluster with three separate email servers setup as containers in a v3.4 cluster. They've worked great for ages. Now, however, I've built a new cluster with v7.1. I was able to restore all the VMs from backups with no problem. The containers, however, are not on the network.

I followed the instructions at https://pve.proxmox.com/wiki/Convert_OpenVZ_to_LXC to make sure I'm converting the container correctly. The end of the tutorial says "voila" but I get "nada." I add a network adapter with an address of 192.168.1.65/24 and try to ping it with no reply. When I connect to the container using "pct enter 1110" (console doesn't work, but I'm sure that's a different problem) and run "ip a" I get:
1: lo: <LOOPBACK> mtu 65536 qdisc noop qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eht0@if14: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop qlen 1000 link/ether 7e:c5:fa:a6:cc:8e brd ff:ff:ff:ff:ff:ff
No listing for the adapter I added from the command line or gui.

Is there a step I'm missing? Or is this some other compatibility issue?

 

[oguz]

hi,

* which distro is running in the containers?

* can you post a config file? pct config 1110

When I connect to the container using "pct enter 1110" (console doesn't work, but I'm sure that's a different problem) and run "ip a" I get:

* does it work after running ip link set eth0 up and/or running dhcp client?

 

[PicklesTheOtt]

hi,

* which distro is running in the containers?

* can you post a config file? pct config 1110

* does it work after running ip link set eth0 up and/or running dhcp clie

The containers are running CentOS 5.8 (vintage stuff).
The config for 1110 is:
arch: i386 cpulimit: 2 cpuunits: 1024 hostname: mail.domain.com memory: 1024 net0: name=eth0,bridge=vmbr0,gw=192.168.1.181,hwaddr=92:68:AE:CA:4B:1A,ip=192.168.1.82/24,type=veth ostype: centos rootfs: Ceph-Storage:vm-1110-disk-0,size=40G swap: 1536

In the container, I ran "ifup eth0" and the network adapter appeared to start. Here's the contents of /etc/sysconfig/network-scripts/ifcfg-eth0:
DEVICE=eth0 ONBOOT=yes UUID=92170944-ebd9-11ec-ac7c-6805ca059c0e BOOTPROTO=none IPADDR=192.168.1.82 NETMASK=255.255.255.0 GATEWAY=192.168.1.181 DNS1=192.168.1.181 DOMAIN=domain.com

But, for some reason, there's no default route. The only routes are:
Destination Gateway Genmask Flags Metric Ref Use Iface 169.254.0.0 * 255.255.0.0 U 0 0 0 eth0 192.168.1.0 * 255.255.255.0 U 0 0 0 eth0

I can manually add a default gateway, but it doesn't retain it on reboot. The network adapter, however, does appear to start properly after a reboot, but again, no default route.

 

[oguz]

The containers are running CentOS 5.8 (vintage stuff)

that's probably the reason for it not working correctly then.
with PVE7 the default was changed from cgroupv1 to cgroupv2 [0]

there are some recommendations/workarounds in the linked section, please take a look

[0]: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#pct_cgroup_compat

 

[PicklesTheOtt]

Just wanted to thank you guys for your help. Moving to proper VMs is the ultimate goal, but in the mean time, I was able to get the containers working in a mostly normal state.

I booted the containers, then logged in using "pct enter <container>" and was able to start the network interface and services manually. After that, when I did a graceful shutdown of the container, it would reboot just fine. If it was stopped, however (either manually or if the node failed), I need to repeat what I did to get it working in the first place. Not a perfect solution, but it let me get everything running again until I can do a full migration.