Cannot remove server from pool :There are still OCFS2 file systems: [Server Pool File System], in the pool. : Oracle VM Manager

So, here's the deal: we need to take that server named "YOUR HOSTNAME" out of the server gang known as "YOUR POOL NAME." This is all about a fancy thing called the Pool FS.

Oh no, we've got a bit of a hiccup. The system is saying "OVMRU_000036E Cannot remove server: 'YOUR HOSTNAME', from pool: 'YOUR POOL NAME' found for discovered Pool FS." Basically, the server isn't budging because there are still some OCFS2 file systems hanging around, particularly one called "Server Pool File System," inside that pool.

Alright, so here's what might be causing this mess. We gotta make sure that the server isn't connected to any repositories when we're trying to give it the boot from the server pool. If that's the issue, the error you're seeing is probably complaining about those lingering OCFS2 file systems.

Let's sort this out, step by step. First, check out that server's repository. Give it a right-click and choose the "present or unpresent" option. This is where you disconnect the server from the pool. Once that's done, go ahead and try to either move the host to an "unassigned" folder or just straight-up remove it from the server pool. If this works, you're golden!

But wait, if it's still throwing a fit...

If there's a pesky pool file system causing a ruckus during the removal process, it might be because other tasks are still messing around with it while we're trying to kick it out. So, give it some time and try the removal again later.

And if the server is being stubborn about accessing its storage and we've done everything we can to calm it down, a simple reboot might do the trick. Sometimes servers just need a moment to get their act together in the cluster before they behave.

Now, here's the heavy hitter: if the server pool's file system has gone haywire or the server's got some leftovers from a past life, it might be time to start fresh. This involves some manual magic on each Oracle VM Server in the cluster. If it comes to this, definitely get in touch with Oracle Support – they're the pros in these situations.

Finally, after a whole lot of searching and experimenting, I stumbled upon a solution that worked for me (found it here: http://blog.whitehorses.nl/2012/05/09/reset-an-oracle-vm-server. Here's the rundown:

1. Log into the server you want to remove using SSH.

2. Stop a service called "ovs-agent" with this command: `service ovs-agent stop`

3. Delete a folder named "db" in "/etc/ovs-agent/" using: `rm -R /etc/ovs-agent/db`

4. Head to "/etc/ocfs2/" and make a backup of "cluster.conf" by copying it: `cp cluster.conf.orig cluster.conf`

5. Start the "ovs-agent" service again: `service ovs-agent start`

Now, give the host machine a good old restart.

Once that's done, open up VM Manager through the web, and go ahead – move that host or kick it out of the pool.

After that, if you're up for it, create a fresh server pool and invite all your host machines back in.

And there you have it – mission accomplished! 🎉