Unable to create pool in Rocky Linux 8 instance using Xen-based instance type in AWS

[vtaneza]

System information

Type	Version/Name
Distribution Name	Rocky Linux
Distribution Version	8.10 (Green Obsidian)
Kernel Version	4.18.0-553.83.1.el8_10.x86_64
Architecture	x86_64
OpenZFS Version	zfs-2.4.0-1

Describe the problem you're observing

The pool creation using zpool create -f <pool name> <device path> failed, seemed to hangup, and after a while, I get kicked out from my ssh session to the EC2 Xen-based instance type m4.large. I tried ssh'ing to the instance again after I get kicked out and I was able to reconnect though but I can't see the pool created using the command zpool list as I get no pools available.

I tried recreating the same steps on a EC2 Nitro-based instance type m5.large and no issues with pool creation at all using the same kernel and zfs version.

Based on the result, it looks like probably not a zfs issue or maybe just some Xen-based block device driver issue and/or incompatibilities with zfs 2.4.

messages.txt
dmesg.txt

Describe how to reproduce the problem

1. Go to AWS console/portal
2. Deploy a Rocky Linux 8 using Xen-based m4.large instance type and attach a 10GB EBS volume into it.
3. SSH to the instance
4. Execute dnf update -y
5. Set the latest kernel installed as default kernel and reboot then install kernel-devel for building zfs
6. Clone, build zfs 2.4 release and install built rpms
7. Create a pool using the volume attached via command zpool create -f <pool name> <device path>
8. After step 7, I observed it took so long for the command to complete, then I get kicked out from my ssh session.
9. I retried steps 1 to 7 but this time, using Nitro-based m5.large instance type and found no issues with pool creation.

Include any warning/errors/backtraces from the system logs

I attached the /var/log/messages and dmesg output if needed.