Proxmox VM is not starting

Question

I have a single-node Proxmox setup and running a few VMs inside it, suddenly today one of the VM stops and is unable to start, I found the below logs from the host node. Any one has any idea
 

oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=ksmtuned.service,mems_allowed=0-1,global_oom,task_memcg=/qemu.slice/1006.scope,task=kvm,pid=1860126,uid=0
kernel: EXT4-fs (dm-38): write access unavailable, skipping orphan clear
cannot delete 'balloon' - not set in current configuration!
 HTB: quantum of class 10001 is big. Consider r2q change.

Answer 1

I have experienced this error in the past, in my case, it was insufficient memory on the host node to allocate to the VM

Comments

Thank you, indeed host node memory was the issue, I have lowered the VM's memory and am able to start now. I will consider upgrading the node resources

Answer 2

The first log message is related to an Out-of-Memory (OOM) killer event, where the kernel is killing a task (with PID 1860126) due to lack of available memory. The OOM killer is configured with a constraint of "CONSTRAINT_NONE", meaning that it can kill any process. The task being killed is a KVM process and is associated with the "ksmtuned.service" cpuset, with allowed memory range of 0-1.

The second log message is related to an issue with an EXT4 file system, where the system is unable to write to the file system due to a lack of write access. The system is attempting to clear the orphan inode table, but it is unable to do so.

The last log message is related to the Linux traffic control (tc) and its Hierarchical Token Bucket (HTB) configuration. The message suggests that the quantum value for class 10001 is large and that the "r2q" value should be considered for adjustment.